Radiology Clinics of North America Cardiac Imaging

Radiol Clin N Am 42 (2004) xi xii
Preface
Cardiac Imaging
Martin J. Lipton, MD
Lawrence M. Boxt, MD
Guest Editors
It is our pleasure to present this issue of the

Radiologic Clinics of North America at a critical time
for cardiovascular radiology. Radiologists have always played a central role in the development of
angiocardiography and nearly all other cardiac diagnostic imaging modalities. However, during the past
quarter of a century, cardiologists have dominated
these procedures. This is due in part to the disproportionate funding of cardiology training grants from
the National Institutes of Health. Furthermore, the
growth of cardiology training programs and cardiologists trained in imaging has increased self referral of
patients for cardiac imaging. This has resulted in the
progressive erosion of cardiac radiology training.
Whereas divisions of cardiac radiology were once
thriving in academic centers, decreased patient volume resulted in decreased stimulus and interest in
learning or teaching in a field that appeared to
exclude radiologist involvement. Today, most radiology training programs have no structured programs in
cardiac imaging, and most radiology departments
have no fellowship-trained cardiac radiology faculty.
In the past decade, CT and MR imaging have
become the dominant diagnostic methods used for
nearly all medical specialties, with the exception of
heart disease. Recent technologic innovation, resulting in high speed electrocardiogram-gated CT and
MR imaging, now allows use of these imaging

modalities for evaluation of the heart. This is the
good news. The bad news is that there is now an
enormous gap between the growth of this new
technology for the diagnosis and management of
patients with heart disease, and the poor knowledge
base obtained by radiologists lacking any clinical
experience in cardiac anatomy, physiology, pathology, and imaging.
This issue attempts to illustrate how this gap can be
narrowed. The plain chest radiograph is discussed by
Drs. Lipton and Boxt in the first article. Plain film
examination has been neglected, yet it provides the
most rapid, cost effective, and safest screening and
diagnostic procedure for identifying and characterizing pulmonary and cardiac pathology. The heart is
displayed in every chest radiograph, which provides a
daily exercise for radiologists in evaluating cardiovascular disease. Furthermore, it is central for understanding the morphologic and physiologic changes
reflected in CT and MR examinations. It is for these
reasons, and because all thoracic and general radiologists feel comfortable with plain films, that this is a
place to begin relearning the basics of cardiac diagnosis. This article emphasizes the importance of a
consistent and systematic approach for analyzing the
heart. It also reminds everyone of the normal cardiac
0033-8389/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.rcl.2004.04.001
xii
M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) xixii
anatomy and geometry of the cardiac silhouette. A

segmental approach is described for the various cardiac borders and their structural relevance. Examples
illustrate how such a logical scheme enables the
observer to reach a differential diagnosis.
This basis is complemented by Dr. Duerinckx in
the next article, which follows and correlates plain
films with MR examination in patients who have
heart disease. It provides excellent illustrations and
examples of classical heart disease findings in the
adult. In the next article, Dr. Boxt describes how
cardiac calcification can be of critical diagnostic
value, which also reinforces the anatomy of cardiac
structures. Dr. Kazerooni and colleagues then describe the postoperative chest and how imaging can
resolve causes of chest pain. The broad spectrum of
postoperative radiographic findings are illustrated by
chest films and CT.
The article by Dr. Woodard describes how to
perform a cardiac MR examination. Now that the
reader is comfortable in understanding the normal
cardiac structures, it will be easier to focus on the
clinical aspects of cardiac MR imaging. Dr. Tatli and
company then describe imaging of thoracic aortic
disease. This is a critical area and is often associated
with emergency room medicine and serious clinical
and medico-legal implications. Detailed MR and CT
methodology is discussed and illustrated for the
practicing clinical radiologist. Pitfalls and key diagnostic points are reviewed.
Dr. Rienmueller and colleagues provide an indepth review of another important and often misunderstood area: the pericardium. The diagnosis of
constrictive pericarditis using CT and MR imaging is
reviewed and illustrated. The important role the
radiologist can play is emphasized with regard to
guiding the surgeons approach and improving patient prognosis.
Dr. Wolfe and colleagues review MR imaging in
ischemic heart disease. The great value of MR in
identifying viable pericardium as well as perfusion
studies are clearly described and illustrated. This area

is one of the fastest growing clinical applications for
cardiac MR. Nuclear imaging including CT/positron
emission tomography is elegantly reviewed and illustrated in the following article by Dr. Coulden
and colleagues.
Dr. Schoepf discusses the technical aspects and
value of CT for examining the coronary arteries,
showing how CT can display not only normal vessel
lumens, but also arterial stenosis and intravascular
stents. In addition, the appearance of plaque progression is demonstrated and discussed. Finally, Dr. Baron
illustrates the value of imaging in congenital heart
disease in adults. Adult patients with congenital heart
disease are becoming increasingly more common as
early diagnosis and surgical and medical management
improves outcome in these individuals.
This issue provides a comprehensive overview of
the contributions radiologists can make in the diagnosis of heart disease. It will also assist the general
radiologist to feel more comfortable in interpreting
and performing noninvasive studies of the thorax.
We are indebted to the very talented individuals
who gave their precious time to prepare their respective articles. We hope that readers will find this issue
enjoyable and useful in their radiology practices.
Finally, our thanks to Barton Dudlick of Elsevier,
who supported this project and exhibited the patience
of a saint.
Martin J. Lipton, MD
Department of Radiology
Brigham and Womens Hospital
75 Francis Street
Boston, MA 02115, USA
Lawrence M. Boxt, MD
Department of Radiology
Beth Israel Medical Center
First Avenue at Sixteenth Street
New York, NY 10003, USA
Radiol Clin N Am 42 (2004) 487 495
How to approach cardiac diagnosis from the

chest radiograph
Martin J. Lipton, MDa,*, Lawrence M. Boxt, MDb
a
Department of Radiology, Brigham and Womens Hospital, Harvard Medical School, 75 Francis Street, Boston, MA 02115, USA
b
Department of Radiology, Beth Israel Medical Center, First Avenue at 16th Street, New York, NY 10003, USA
Direct visualization of pathologic changes in the

heart has significantly influenced the diagnostic accuracy and management of patients with cardiovascular
disease. Fluoroscopic and plain film examination
quickly became important early tools for the diagnosis
of heart disease [1]. Angiocardiography ushered in an
era of surgical management of acquired and congenital heart disease, and also became crucial for guiding
percutaneous catheter-based therapeutic interventions.
Radiologists played a central role in the development
of these imaging modalities, and as such enjoyed
significant participation in the diagnosis and management of patients with heart disease.
Growth in the clinical use of cardiac MR imaging
and CT not only reflects the noninvasive importance
of these technologies, but also is an expression of
societal issues concerning the risks and costs of
providing cardiac diagnosis and management of these
patients. Again, radiologists influenced the development and evolution of these technologies, which is a
reflection of the interests and expertise of the international radiology community. In this era of growing
interest in noninvasive cardiovascular diagnosis,
however, there exists a significant gap between the
growth in radiologic technologies for the diagnosis
and management of patients with heart disease, and
the instruction and early clinical experience that
radiologists obtain during their residency training.
Despite the important contribution of imaging to
patient diagnosis and management, and the expanding role of the newer modalities in the management of
* Corresponding author.
E-mail address: mlipton@partners.org (M.J. Lipton).
these patients, instruction in the radiographic evaluation of the heart is limited or totally neglected in
many training programs. In particular, plain film
evaluation of the heart is trivialized, and radiology
residents are not instructed in their interpretation. A
dangerous gap exists between the use of imaging
studies and the ability of radiologists to perform and
interpret these examinations.
Plain film examination is still the most commonly
performed cardiac imaging test obtained in the United
States. The use of plain film examination has evolved
from being the only and most valuable imaging tool,
to its current use for detection of cardiac chamber
abnormalities and evaluating the instant state of
cardiac physiology [2]. Along with this evolution,
apparent institutional interest in cardiac plain film
examination has waned. Academic medical centers
fail to train radiology residents in the plain film
examination of the heart. By emphasizing new,
high-technology imaging modalities, notably MR
imaging and multidetector spiral CT, residents and
practicing radiologists are not availed of the sensitivity and reliability of the chest film for evaluating a
patients cardiac status. Reduced interest and awareness of radiologists limits the value of their interpretation, leading not only to underuse but also to
distrust of reported findings. This is sad and unfortunate, because plain film cardiac examination is
perhaps the most rapid, cost-effective, and safest
screening procedure for identifying and characterizing pulmonary and cardiac pathology. The chest
radiograph examination displays the heart in every
individual examined, and provides a daily exercise in
evaluating cardiovascular disease that is central for
understanding the morphologic and physiologic
doi:10.1016/j.rcl.2004.03.009
488
M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495
changes reflected in CT and MR imaging examinations [3]. The chest film is so closely tied to instruction in and understanding of disease mechanisms and
morphologic changes that result from these mechanisms, that the abandonment of training in plain film
diagnosis undermines the ability to maintain expertise
and control of CT and MR imaging of the heart. If
residents are not trained in plain film diagnosis of
heart disease, they will not have the tools to compete
for the use of cardiac CT and MR imaging. Nevertheless, so long as interpretation of the chest radiograph remains primarily the responsibility of the
radiologist, they are responsible for training radiologists competently to perform this service. To this end,
this article presents an approach to plain film examination of the heart. The approach is based on basic
principles of radiologic evaluation. It emphasizes the
relationship between the radiologic appearance of a
structure and the technique used to obtain that image,
and the relationship between the observation of a
structural abnormality and the anatomic relationships
that allow that observation to be made. This approach
is simple, organized in a logical manner, and when
applied rigorously results in not only accurate and
insightful differential diagnosis, but also a deep
understanding of cardiovascular disease processes,
which are essential for the best use of CT and
MR imaging examinations of the heart.
Approach to the film

There are two fundamental requirements for the
radiologist when interpreting a chest radiograph from
the cardiac diagnostic perspective. The first is to
examine the image in a consistent and systematic
manner. One systematic approach is given in Box 1.
Box 1. A cardiac diagnostic approach to

the chest radiograph
Technical factors
Skeletal abnormalities and hardware
Situs: gastric air bubble, cardiac
apex, and aortic knob
Box 2. Frontal chest radiograph: normal

border-forming structures
Left upper border
Left subclavian artery, aortic arch
(aortic knob)
Left middle border
Pulmonary trunk (main pulmonary
artery), left arterial appendix
Left lower border
Left ventricle
Right upper border
Innominate vein and superior vena
cava or innominate artery and
ascending aorta
Right lower border
Right atrial appendix and right atrium,
inferior vena cava
This approach serves well in developing a disciplined

and logical scheme.
Radiologic technique
Before interpretation commences, the first step is
to evaluate the quality of the examination itself. Not
only should the observer be interested in the size and
shape of the radiologic contours, but also the radiologic technique used to obtain the image. Estimation
of the radiographic technique helps the observer to
assess the severity of pulmonary vascular changes. It
should be possible to visualize the thoracic spine
through the mediastinal shadow. Radiographic underexposure (light films) generally results in overestimation of the unsharpness of vessels leading to
overinterpretation of pulmonary vascular congestion.
Conversely, overpenetration (dark films) may produce better visualization of the mediastinal structures,
but prevents detailed observation of the pulmonary
parenchyma, leading to underestimation of congestion or the feeling that pulmonary blood flow is diminished. Equally important is an estimation of the
degree of inspiration, because this changes the appearance of the heart size and the pulmonary vessels.
Heart: position, size, and shape

Great vessels: position, size, and
shape
Lung fields and vascularity by zone
Search for calcifications
Patient position
Patient rotation changes the borders of the heart,
bringing some structures into profile, and others out.
Box 3. Lateral chest radiograph: normal

border-forming structures
Radiologic anatomy and heart size
Anterior lower border
Right ventricle
Posterior upper border
Left atrium and pulmonary veins
Posterior lower border
Left ventricle (sometimes right atrium),
inferior vena cava
The ability to diagnose is based on the expectation of

the heart presented in certain projections. Patient
rotation moves expected heart border-forming structures away from the border. Evaluation of the relationship of the trachea, the clavicular heads, and the
sternum, the symmetry between right and left ribs,
and positions of the humoral heads are good indicators of the rotation of the torso in the examination.
The trachea should be in the midline. The distance
between the clavicular heads and the sternum and the
size of the left and right hemithoraces should be
equal. The curving of the left and right ribs should
be symmetric. To the extent that any of these criteria
are not met, the chest is rotated, and awareness of this
variance plays a part in the evaluation of observed
morphologic changes [4].
489
Radiographic projection
Conventional chest film examination of the heart
and lungs is performed using a six-foot posteroanterior beam. An upright patient faces the film screen
combination, and the radiographic exposure comes
from behind (the view is named by the course of the
beam, hence posteroanterior). Using this method,
the X-ray beam is nearly parallel (nondiverging).
The ventrally located heart is close to the film-screen,
and is not (minimally) magnified. Emergency room
or intensive care chest radiography is typically
performed from in front of the patient who leans
against the film cassette. In this anteroposterior
radiograph, the heart is away from the film-screen
combination, and the X-ray beam is not parallel and
diverges. These factors result in magnification of
the heart.
Chest radiographs obtained with the patient upright provide a reliable representation of the distribution of interstitial lung water and pulmonary blood
flow. The lower pulmonary lobes contain more parenchyma and receive more blood flow than the
upper lobes. The lower lobe pulmonary arteries and
veins should be greater in caliber than the upper lobe
branches. When supine, the lower lobe pulmonary
vessels lose their gravity dependence, and the upper
lobe pulmonary vessels become dependent. In the
supine chest radiograph, the upper lobe vessels appear greater in caliber than the lower lobe vessels,
giving the lungs the appearance of pulmonary redis-
Fig. 1. (A) Frontal chest radiograph, normal (expected) anatomy. (B) Diagram of border-forming structures, frontal view (From
Jefferson K, Rees S. Clinical cardiac radiology. 2nd edition. Butterworth; 1980; with permission.).
490
tribution, even in an individual with normal cardiac function.
Skeletal abnormalities
Skeletal abnormalities are important observations
and often are only observed by a directed visual
search of the bony thorax, including the identification
of rib notching, sternal depression, vertebral body
erosion, premature sternal fusion, and scoliosis. Situs
abnormalities are of particular importance when
congenital heart disease is present or suspected and
may coexist.
Examples of abnormalities of specific segments
of the cardiac silhouette are used to describe how a
cardiac diagnosis is reached; these are listed in
Boxes 2 and 3. An organized and disciplined visual
search is the first fundamental requirement.
The second fundamental requirement for the radiologist is an understanding of the basic cardiac
radiographic anatomy. This knowledge is critical
and is currently poorly taught. It is essential for
developing a logical process of deduction, which
guides the search for a constellation of observations
necessary to deduce a differential diagnosis. The
normal border-forming structures in the mediastinum,
which determine the cardiovascular silhouette in the
Fig. 3. Coronal scan through chest obtained with a spin echo

MR imaging sequence demonstrating the major vascular
structures. Note that the diameter of the normal aorta and
main pulmonary artery are approximately equal in size.
frontal projection, are bounded by radiolucent lung

and are listed in Box 2, which the reader should
compare with Fig. 1.
The lateral chest radiograph is depicted in Figs. 2, 3
and this correlates with the normal border-forming
structures listed in Box 3.
Fig. 2. (A) Normal lateral chest radiograph. (B) Diagram of border-forming structures in the lateral projection (From Jefferson K,
Rees S. Clinical cardiac radiology. 2nd edition. Butterworth; 1980; with permission.).
Box 4. Questions to ask

Where is the aortic knob?
Is it normal in size, shape, and
position?
How does its size compare with the
main pulmonary artery? Normally

they should be approximately the
same size (Fig. 3).
Heart size
The size of the cardiac silhouette has importance,
because it may represent several underlying disease
processes [2]. It may be evaluated subjectively, or by
measuring the cardiothoracic ratio or by volume
measurement. Subjective assessment is the most
common method used by the experienced observer.
Technical factors mentioned previously should always be taken into consideration.
Aortic knob
It is surprising how often findings involving this
segment are overlooked. Box 4 states the issues very
simply. If the aortic knob cannot be identified,
congenital abnormalities should be considered, including a right-sided arch, coarctation of the aorta, or
491
double aortic arch. A barium swallow should be

obtained in both frontal and lateral projections to
identify an aberrant left subclavian artery as shown in
Fig. 4. In the presence of a right arch the trachea is
deviated to the left, and this sign may be very helpful.
A smooth posterior impression on the barium column
in the lateral projection usually indicates an aberrant
subclavian artery and denotes the presence of a
vascular ring, which is commonly asymptomatic.
The absence of the aberrant vessel usually indicates
mirror-image branching of the great vessels and is
almost always associated with other cardiac malformations, most commonly tetralogy of Fallot, truncus
arteriosus, transposition, and ventricular septal defect.
The plain film may offer additional help if the aorta
can be identified descending on the left side, in
which case the heart is usually normal; if it descends
on the right side, then congenital heart disease is
usually present.
Cross-sectional imaging with contrast-enhanced
CT or MR imaging provides a definitive diagnosis.
A limited barium esophagram, however, is far less
expensive and simpler to obtain. Furthermore, a good
technologist can frequently recognize a questionable
right arch and either gives barium while the patient
is still in the chest room or at least checks with
a radiologist.
This article does not describe exhaustive lists of
differential diagnoses. The examples chosen serve the
purpose of emphasizing the value of a disciplined and
directed visual search for specific cardiac structures
Fig. 4. (A) Right-sided aortic arch demonstrated on a frontal chest radiograph during a barium swallow. (B) Lateral radiograph
in the same patient shown in 4A. Note the smooth filling defect posteriorly on the barium near the level of the aortic arch caused
by an aberrant left subclavian artery. This finding indicates some form of vascular ring and probably no serious congenital
heart disease.
492
Fig. 5. (A) Chest radiograph obtained in a young woman as part of a routine physical. Note that the aortic knob is abnormal
in size and appearance with a rim of calcification laterally. The remainder of the image is normal. (B) An aortogram in the
same patient as shown in 5A in a frontal projection demonstrating a calcified false aneurysm of the aorta, which was a sequel to
aortic arch transection 18 years earlier caused by trauma from an automobile accident.
in interpreting routine chest studies. See the article by

Duerinckx elsewhere in this issue for further exploration of this issue.
Fig. 5 illustrates another patient in whom this
chest radiograph was obtained during a routine physical for a new job. Examine this radiograph and try to
identify an abnormality before reading this text further or the legend of Fig. 5A. The patient, on detailed
specific questioning, remembered being a passenger
in a car crash 18 years earlier, but saw no physician
Fig. 6. This radiograph illustrates the classical findings of

localized enlargement of the main pulmonary artery in a
teenager. The right and other distal branches of the
pulmonary artery look normal. The diagnosis is pulmonic
valve stenosis with poststenotic dilatation.
and has always been asymptomatic. The aortic knob

is enlarged, has a mixed density, and a rim of
bordering calcification. Although the differential diagnosis includes other lesions, this represents a calcified pseudoaneurysm of the arch caused by a
chronic traumatic transection of the aorta. Fig. 5B
shows an aortogram in the frontal plane, which con-
Fig. 7. Frontal chest radiograph demonstrating enlargement

of the main pulmonary artery segment, and also grossly
enlarged proximal and secondary and tertiary branches of
the right pulmonary artery. The cardiac apex is elevated with
some straightening of the left cardiac border; these findings
indicate a diagnosis of pulmonary arterial hypertension.
Box 5. Causes of pulmonary arterial

hypertension

Destructive lung disease

Mitral valve stenosis
Pulmonary embolism
Intracardiac shunts: congenital or
iatrogenic
Extracardiac shunts (eg, A-V fistulae)
Constrictive pericarditis
Idiopathic
firms the finding. This patient is at risk of rupture

even after all these years and requires surgical repair.
Pulmonary artery
As noted previously, the normal aortic knob and
normal main pulmonary should be approximately
equal in size. This observation must be based primarily on examining that portion of the arc of each great
artery, which is visible on the frontal radiograph.
Fig. 6 demonstrates an asymptomatic young patient
in whom the only finding is an enlarged main
pulmonary artery segment. This is too large to be
simply physiologic as occurs frequently in young
women under 30 years. The important observations
lie in analyzing the right and left proximal pulmonary arteries, which in this patient are normal. This
key observation excludes the diagnosis of pulmonary
493
arterial hypertension, and indicates pulmonic valve

stenosis with poststenotic dilatation.
Compare the radiograph in Fig. 6 with that of
another patient illustrated in Fig. 7. Note that the
main pulmonary artery segment here is grossly enlarged compared with the aortic knob, as it also was
in Fig. 6. Look at the right hilar vessels, however, and
notice how they too are enormous. Enlargement of
the tertiary pulmonary vessels as depicted in Fig. 7 is
never seen in isolated pulmonic valve stenosis. This
distinction can and should be made by the radiologist
on a routine basis. The causes of pulmonary arterial
hypertension are listed in Box 5. Many of these listed
disorders can be excluded or considered likely by
further examining the fronted and lateral chest images
in any given patient. Mitral valve stenosis is near the
top of the list of diagnoses given in Box 5, because it
can be treated and often cured by modern surgery,
hence the importance of not forgetting to consider it
as a possibility. Idiopathic pulmonary arterial hypertension is at the end of the list for the opposite reason,
because the treatment is limited and primarily aimed
at palliation. Pericardial constriction is discussed in
detail in the article by Reinmuller et al elsewhere in
this issue.
Left atrium
The segment of the left atrial appendage should
not be convex outward from the heart. When it is
enlarged it usually indicates enlargement of the body
of the left atrium. Fig. 8A is an example of a localized
Fig. 8. (A) Chest radiograph showing an abnormal contour in the region of the left atrial appendage, which is usually associated
with an enlarged left atrium. (B) Lateral view of the same patient shown in 8A demonstrating a barium-filled esophagus, which
is displaced by the enlarged left atrial chamber.
494
Cardiac valve disease
Fig. 9. Diagram of a lateral chest radiograph indicating the

findings of right ventricular and left atrial enlargement,
typical in patients with mitral stenosis.
enlargement of the left atrial appendage along the left

cardiac border. The left atrium lies in a rather high
position on the lateral projection, as seen in Figs. 2B
and 8B. This point is also shown diagrammatically in
Fig. 9. In mitral valve stenosis the left atrium and
right ventricle are typically both enlarged. In addition, blood diversion to the upper lobes in the upright
posture is the third characteristic sign of mitral valve
stenosis, along with enlargement of the left atrium
and right ventricle.
The normal position of each of the four heart

valves should be known. This is important when
deciding the location of valve calcification or when
a valve prosthesis is present. Fig. 10 demonstrates a
patient in whom three valves were replaced by balland-cage prosthesies. Note that the mitral and aortic
valves are in continuity, whereas the tricuspid, the
most inferior valve, is separated significantly from
the pulmonary valve, which is the most superior
valve and has in this case not been replaced. See
the article by Baron elsewhere in this issue for further
exploration of this topic.
The direction of the cage determines the direction
of blood flow and helps to decide which chambers
are involved in the valve in question. Additionally, a
line can be drawn from the left atrial appendage to the
point of intersection of the right atrium and diaphragm. The aortic valve lies above and the mitral
below this line. A similar imaginary line can be
drawn for the lateral chest radiograph as illustrated
in Fig. 10B. This line is drawn from the carina to the
point where the sternum is intersected by the left
diaphragm [5].
Duerinckx shows excellent examples of other
plain film cardiovascular findings elsewhere in this
issue. He correlates MR images with the plain chest
radiograph in patients with left ventricular aneurysm
and aortic aneurysm. He also shows examples of
Fig. 10. (A) Triple valve replacement and generalized cardiomegaly. Braunwald-Cutter prostheses in the aortic (A), mitral (M),
and tricuspid (T) positions. All four chambers are enlarged with left atrial enlargement suggested by splaying of the carina
(arrows). (B) Lateral radiograph in the same patient. Aortic prosthesis lays anterosuperior to mitral. Enlargement of right-sided
chambers is indicated by filling in the anterior mediastinal window (From Coulden R, Lipton MJ. Radiological examination in
valvular heart disease. In: Al Zaibag M, Duran CMG, editors. Valvular heart disease. New York: Marcel Dekker; 1994. p. 162;
with permission.).
495
Summary
Fig. 11. The heart lies almost entirely within the left
hemithorax, yet the patient is not rotated. In the absence of a
reduced anteroposterior distance this appearance suggests
complete congenital absence of the left pericardium.
Sometimes, as in this patient, the aorta and main pulmonary
artery are more sharply defined than normally, because of
the presence of lung tissue lying between the great vessels.
aortic valve stenosis, Marfan syndrome, and right

heart abnormalities including Ebsteins disease. Baron
shows examples of congenital heart disease elsewhere in this issue and these are not duplicated in
this section.
Position of the heart

Normally approximately one third of the heart lies
to the right and two thirds to the left of the midline.
When it is displaced as in Fig. 11 into the left
hemithorax, this finding must be explained. The
commonest cause is rotation of the patient to the left.
Another cause is a narrow anteroposterior diameter
caused for example by a depressed sternum. When
these two explanations are not present the possibility
of complete absence of the left pericardium should be
entertained. An MR imaging study confirms this rare
congenital abnormality. This case serves to emphasize the importance of this key observation (ie,
recognizing that the heart is displaced in an otherwise
normal chest radiograph), because it makes the diagnosis. Position is as important as size and shape when
examining the whole heart or any border-forming
cardiovascular segment.
Traditional plain chest radiography provides the

earliest opportunity in many instances for diagnosing
all forms of heart disease. Indeed, it is perhaps the
most rapid, cost-effective, and safest screening procedure for identifying lung pathology. Its role in heart
disease has been undervalued, however, and academic
centers have simply not provided adequate training in
cardiac imaging. This may be understandable given
the progressive trend during the past three decades, in
which cardiac imaging has become primarily the
domain of cardiologists. Interpretation of the chest
radiograph remains primarily the responsibility of
the radiologist. The most significant limitation of the
chest radiograph as a cardiac diagnostic tool is the
experience and knowledge of the physician who
interprets the examination.
Two fundamental concepts were emphasized. The
first is a requirement for every radiologist to develop
a systematic and consistent approach for analyzing
frontal and lateral chest radiographs to identify any
cardiac and vascular abnormalities. This visual search
pattern may vary for different radiologists, but whatever system is adopted it should be applied consistently and routinely to include evaluating the areas
listed in Box 1.
The second fundamental concept is the need to
have a working knowledge regarding the location and
normal size and shape of the various segments
bordering the cardiac silhouette. This is the basis on
which the radiologist recognizes specific abnormalities and associates them with a meaningful differential diagnosis.
References
[1] Jefferson K, Rees S. Clinical cardiac radiology. 2nd
edition. Butterworths; 1980.
[2] Lipton MJ. Plain film diagnosis of heart disease: cardiac
enlargement. Contemporary Diagnostic Radiology 1988;
11:1 6.
[3] Boxt LM, Reagon K, Katz J. Normal plain film examination of the heart and great arteries in the adult.
J Thorac Imaging 1994;9:208 18.
[4] Boxt L. Plain film examination of the normal heart.
Semin Roentgenol 1999;34:169 80.
[5] Coulden R, Lipton MJ. Radiological examination in valvular heart disease. In: Zaibag MA, Duran C, editors.
Valvular heart disease. New York: Marcel Dekker;
1994. p. 131 83.
Radiol Clin N Am 42 (2004) 497 514
How to plan and perform a cardiac

MR imaging examination
Mehdi Poustchi-Amin, MDa, Fernando R. Gutierrez, MDa,
Jeffrey J. Brown, MDa, Scott A. Mirowitz, MDb, Vamsidhar R. Narra, MDb,
Naoki Takahashi, MDc, Gary R. McNeal, MSd, Pamela K. Woodard, MDa,*
a
Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 South Kingshighway Boulevard,
St. Louis, MO 63110, USA
b
Department of Radiology, University of Pittsburgh Medical Center, Pittsburgh, PA, USA
c
Kyushu University, Japan
d
Siemens Medical Solutions USA, Inc., 51 Valley Stream Parkway, Malvern, PA 19355, USA
Cardiovascular disease is the major cause of death

in the Western world. It absorbs a large portion of the
United States health care budget and has both social
and economic consequences. Despite many wellestablished techniques for the diagnosis of cardiac
disease, a full diagnostic work-up may require examinations with several different imaging modalities.
This is expensive and time consuming. MR imaging
of the heart has the potential to provide information
often supplied by a number of modalities in a single
examination. Cardiac MR imaging provides anatomic
and functional diagnosis of acquired and congenital
heart disease. Moreover, it is a precise technique for
the quantification of ventricular dimension and function [1], and more recently has been used to assess
myocardial viability and perfusion [2,3]. With recent
technical advances in the field of cardiovascular MR
imaging, there will be an increased demand for
clinical cardiac MR imaging. This article provides
the reader with a basic understanding of cardiac MR
imaging and the practical applications required to
perform cardiac MR imaging.
This article is a revision of an article published in Magn

Reson Imaging Clin North Am 2003;11:1 18.
E-mail address: woodardp@mir.wustl.edu
(P.K. Woodard).
Cardiac MR imaging techniques: general

principles
Before describing the general principles of cardiac
MR imaging, it should be mentioned that, unlike
imaging of other organ systems, the protocol performed is highly dependent on the question to be
answered. The protocol for assessing cardiac ischemia differs greatly from the protocol for assessing
coronary arteries or congenital heart disease. The
specific reason for the performance of the examination should be determined, permitting the physician
to tailor the examination to a given question. Communication with the patient is also important. First,
the patient should be screened for any contraindications to the MR imaging examination. This routinely
includes queries regarding the presence of pacemakers, ferromagnetic implants, or intracranial aneurysm clips. If pharmacologic myocardial stress agents,
such as adenosine or dobutamine, are included in
the protocol, the patient should also be queried about
relative contraindications to these agents and instructed not to eat before the examination. For instance, contraindications to adenosine include active
bronchospasm, first-degree heart block, or systolic
blood pressure of less than 90 mm Hg. In addition,
patients to received adenosine should be informed to
refrain from partaking in any caffeine, chocolate, or
theophylline-containing drugs during the 24-hour
period before the MR imaging examination. They
doi:10.1016/j.rcl.2004.03.004
498
M. Poustchi-Amin et al / Radiol Clin N Am 42 (2004) 497514
should be instructed that they might feel flushed or

nauseated. Instructing all patients in breathholding
techniques before patient entrance into the scanner
is an efficient method of improving image quality.
Patients should also be informed that motion during
scanning could limit the quality of the examination.
Cardiac gating and physiologic monitoring

In most instances, to avoid image blur, MR imaging acquisition must be limited to a constant portion
of the cardiac cycle. This is accomplished through
cardiac gating. Gating can be triggered to the QRS
complex of the ECG, or if that fails to a peripheral
pulse. This coordination of imaging to cardiac contraction reduces flow and motion artifacts. The most
effective gating involves the use of the ECG signal
(ECG gating). The objective of ECG gating is to
acquire an R wave that is substantially larger than
the T or S wave of the ECG. The most common
configuration for placement of ECG leads for cardiac gating is shown in Fig. 1. It is important
to obtain good skin contact by shaving the skin if
necessary and cleansing the skin, preferably with
commercially available abrasive agents, and by the
use of ECG electrodes with coupling gels. Even after
obtaining a strong R wave before the examination,
once the patient enters the magnet or after MR
imaging has started, additional noise caused by the
magnetic field and the radiofrequency pulse may

obscure the R wave. Repositioning the electrodes to
obtain a better ECG signal and the use of newer
fiberoptic leads to reduce interference may help to
reduce some of these problems.
In addition to cardiac gating, physiologic monitoring is necessary during many types of cardiac MR
imaging examinations. This is especially true if the
examination is being performed for assessment of
ischemic cardiac disease, when adenosine or dobutamine is administered, or when anesthesia or sedation
is administered for pediatric patients. MR imaging
compatible hemodynamic monitoring systems are
now available permitting continuous monitoring of
pulse, blood pressure, oxygenation, and heart rhythm
throughout the MR imaging examination.
Cardiac MR imaging pulse sequences

With recent technologic advances in MR imaging
hardware and software there are now many pulse
sequences available for cardiac MR imaging. Pulse
sequences currently used for cardiac imaging can
be generally divided into dark-blood and brightblood techniques. In dark-blood or black-blood
techniques, fast-flowing blood is black or of low
signal intensity. These techniques produce images
for anatomic delineation of blood vessel lumen and
cardiac chambers [4,5]. Examples of this technique
Fig. 1. Cardiac gating: the most commonly used configuration for ECG lead placement.
include conventional spin echo, breathhold turbo or

fast spin echo (TSE, FSE), and half-Fourier turbo
spin echo sequences with double inversion recovery
(IR) pulses to suppress blood signal (HASTE, double-IR TSE-FSE).
Generally speaking, for dark-blood sequences the
effective (or total) TR should be approximately 85%
to 90% of (or 100 millisecond less than) the patients
R-R interval (time between R waves). For instance,
the actual TR plus any trigger delay, if necessary, is
approximately 85% to 90% of the R-R interval. Note
that for T1-weighting, TR should be less than
900 milliseconds. For double IR sequences, which
are dark-blood but T2-weighted, the TR should
remain long and the acquisition window should cover
two heart beats [6,7].
In bright-blood techniques, flowing blood is white
or of high signal intensity. These are routinely gradient recalled echo sequences (GRE). Cine GRE
sequences that produce a motion picture loop throughout the various phases of the cardiac cycle are particularly useful. GRE images can be obtained with
segmented k-space technique and cardiac gating. A
single-slice multiphase or multislice single cardiac
phase mode can be performed in a short breathhold
period. Examples for various vendors include fast low
angled shot (TurboFLASH), fast spoiled gradient
recalled echo, and turbo field echo and fast field echo.
The parameters for these sequences are adjusted to the
patients breathholding capability and heart rate. For
patients with slower heart rates, sequences that provide a greater number of lines per segment can help to
shorten the required breathhold. The number of
phases of the cardiac study should be set according
to the following formula [8,9]:
Number of cardiac phases R R interval
85%=TRef f ective
Newer fast, short TE GRE sequences with completely refocused gradients provide excellent contrast
between the myocardium and blood pool and are
commercially known as trueFISP, balanced fast field
echo, or FIESTA.
Various types of MR imaging pulse sequences
provide different information. It is important to know
that functional abnormalities may not be examined
directly by dark-blood techniques, but only inferred
by analysis of resultant morphologic changes [8]. For
example, aortic regurgitation can be inferred from
the findings of an enlarged left ventricle and dilated
ascending aorta. Cine bright-blood technique with
high temporal resolution allows functional analysis,
however, including demonstration of the regurgitant
499
jet and quantification of aortic regurgitation. Most

cine acquisitions have lower contrast resolution resulting from the short flip angle and short TR used.
The exceptions are the completely refocused GRE
sequences, which provide excellent contrast-to-noise
ratio. As a result, these trueFISP-type sequences are
very useful for segmenting the myocardium from the
blood pool and are excellent for functional assessment of the myocardium. Nevertheless, it should be
remembered that because the TE is so short in these
sequences, less dephasing occurs, decreasing the
visibility of stenotic or regurgitant jets. In addition,
these sequences are less useful than standard GRE
cine sequences for the imaging of valve leaflets [9].
Standard GRE sequences should be used in the
assessment of cardiac valves, or when attempting to
identify intracardiac shunts including atrial septal or
ventricular septal defects. As a general rule, imaging
should begin with dark-blood sequences to obtain
anatomic information and proceed with bright-blood
techniques to assess functional abnormalities.
Cardiac imaging planes

The planes generally used for imaging the thorax
are the three orthogonal planes of the thorax (transverse, sagittal, and coronal) with the patient supine.
Because the cardiac axes are not parallel to the body
axes, however, sections parallel and orthogonal to
cardiac axes (short axis and long axis of the heart) are
often favored for cardiac imaging [10 12]. These
have the advantage of generally corresponding to the
planes used with other noninvasive cardiac imaging
modalities. A phase array surface coil or dedicated
cardiac coil is necessary to obtain a good signal-tonoise ratio. The examination usually begins with a
general anatomic survey using a dark-blood technique in one or more of the three planes: axial,
coronal, and sagittal.
Scout images: transverse or axial plane
The first plane that may be obtained is an axial
survey of the chest. This is the imaging plane most
familiar to the general radiologist. Most anatomic
structures are easy to identify on this plane and the
overview permits assessment of adjacent thoracic
pathology. Transverse or axial images (Fig. 2) at the
base of the heart display the normal relationships of
the great vessels and cardiac chambers. Portions of
the proximal coronary arteries near their origin and
pericardium can also be displayed. Axial sections are
especially useful in the evaluation of congenital heart
500
Fig. 2. Transverse or transaxial images: (A) Dark-blood technique: single-slice breathhold turbo spin echo (TSE) T1 is often used
to assess cardiac morphology. (B) Bright-blood technique: breathhold cine gradient recalled echo (GRE) sequence is useful in
assessing cardiac function. LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle.
lesions and may complement morphologic evaluation

of patients with acquired heart disease.
Coronal and sagittal planes

For anatomic imaging, coronal and sagittal planes
can also be acquired. The coronal plane (Fig. 3)
is often effective for demonstrating the aortic valve.
More posteriorly, coronal planes show the entrance
of the upper lobe pulmonary veins into the left
atrium. It is also useful for showing the diaphragmatic surface of the left ventricle and the extension
of pericardium over the proximal portion of the
great arteries.
Double-oblique (oblique-sagittal) planes through

the pulmonary trunk and aorta (Fig. 4) are useful for
demonstrating the pulmonic and aortic valves and
outflow tracts. Other anatomy well seen on doubleoblique images includes the connections of the superior and inferior vena cavae to right atrium, and one
or more sinuses of Valsalva. The plane parallel to the
axis of aortic arch, seen on axial images, is used to
obtain oblique-sagittal images for evaluation of aortic dissection.
After obtaining any desired orthogonal views,
many cardiac MR imaging studies require images
parallel to the true short and long axis of the heart.
Because the heart lies obliquely in the thoracic cavity,
the true long axis of the heart is oriented approximately
Fig. 3. Coronal images. (A) Dark-blood technique: TSE T1. This plane nicely demonstrates the aortic valve (arrow). A plane set
through the mid aortic valve and left ventricular apex provides a five-chambered view. (B) Bright-blood technique: cine GRE.
This plane can be used to assess the jet of aortic stenosis or insufficiency. Ao, aorta; LV, left ventricle; RA, right atrium; RV,
right ventricle.
501
Fig. 4. Bright-blood double-oblique images through pulmonary trunk (A) and aorta (B). Ao, aorta; LV, left ventricle; PT,
pulmonary trunk; RV, right ventricle.
45 degrees to the mid-sagittal plane of the thoracic

spine. These short- and long-axis views of the heart are
preferred for quantification of ventricular dimensions
and regional contractile function [10 12]. Because
similar views are obtained during the echocardiogram,
these planes are often familiar to cardiologists.
Vertical long-axis plane (two-chamber view)
The vertical long-axis plane or two-chamber view
(Fig. 5) is used to evaluate the left heart structures. It
reveals information concerning superoinferior and
anteroposterior anatomic relationships and is useful
for assessing the mitral valve. This plane is prescribed
from an axial image, which shows the largest oblique
diameter of left ventricle.
Horizontal long-axis plane (four-chamber view)
Images prescribed from the left ventricular long
axis (two-chamber view), set up through the posterior
wall of the left atrium, mitral valve, and left ventricu-
lar apex, provide a horizontal long-axis or fourchamber view of the heart (Fig. 6). The horizontal
long-axis plane or four-chamber view displays the
relationship of the four cardiac chambers to each
other on a single image. Cine GRE images obtained
in this plane display mitral, tricuspid, and aortic valve
function and right and left ventricular contraction.
This image plane can also be obtained by oblique
transverse imaging through a short-axis scout.
Short-axis plane
The short-axis plane (Fig. 7) is obtained when
images are prescribed perpendicular to left ventricular
long axis seen on a two-chamber view. It shows the
true cross-sectional dimensions of cardiac chambers.
Initial images in this plane are performed through the
papillary muscles, with subsequent images performed
toward the heart apex and base. In this plane the left
ventricular myocardium is displayed as a doughnutshaped ring. Cine GRE images allow visualization
and quantification of systolic myocardial wall thick-
502
Fig. 5. Two-chambered long-axis view. An image obtained parallel to the line shown on the transaxial image (A) provides the
vertical long-axis plane or two-chamber view (B). This image plane is ideal for assessing the mitral valve. LA, left atrium; LV,
left ventricle.
ening. This plane can also be used for quantifying left

and right ventricular volume and mass and ventricular
ejection fraction when the appropriate software is
available. Differences between right and left ventricular stroke volumes can be used to estimate valvular
regurgitation or shunt ratios.
Long-axis view through aortic and mitral valves

This view, obtained through the left ventricular
apex and aortic outflow tract, is prescribed from a
coronal image (Fig. 8). This plane demonstrates both
the aortic and mitral valves. Because it displays por-
Fig. 6. Horizontal long-axis plane or four-chamber view (GRE, TrueFISP). An image obtained parallel to the line shown on the
vertical long-axis image (A) provides the horizontal long-axis or four-chamber view (B). In this image, both the mitral and
tricuspid valves can be assessed. LV, left ventricle; MV, mitral valve; RV, right ventricle; TV, tricuspid valve.
503
Fig. 7. Short-axis plane. Bright-blood technique cine GRE. This image plane is favored in the assessment of left ventricular
function. Multiple contiguous short-axis images are obtained from the base of the heart to the apex (A) to provide images in the
short-axis orientation (B). Functional analysis software can then be used to calculate stroke volume, ejection fraction, and
myocardial mass. A horizontal four-chamber view can be prescribed from a short-axis image by drawing a line perpendicular to
the left ventricular septum. LV, left ventricle; RV, right ventricle.
tions of the left ventricle, right ventricle, left atrium,

right atrium, and ascending aorta, it is sometimes
known as the five-chamber view.
Routine clinical studies with cardiac MR imaging

Reasons for a cardiac MR imaging examination
most frequently include preoperative and postopera-
tive congenital heart disease assessment; clinical

suspicion of right ventricular dysplasia; pericardial
disease (constrictive pericarditis versus restrictive cardiomyopathy); cardiac tumors; anomalous coronary
arteries; and valvular disease. In addition to the indications described previously, MR imaging can now
be used to assess myocardial function and viability.
Sequences also are being investigated to assess for
proximal coronary artery stenoses and myocardial
Fig. 8. A line drawn through the left ventricular apex and aortic outflow as prescribed from a coronal image (A) provides a longaxis view sometimes known as the five-chamber view (B). This view demonstrates both aortic valve and mitral valve function
and displays portions of the right and left ventricles and atria and the aorta (five chambers). Ao, aorta; LA, left atrium; LV,
left ventricle; RA, right atrium; RV, right ventricle.
504
perfusion. The following section discusses and illustrates each indication, MR imaging techniques including pulse sequences, and some clinical cases.
Adult congenital heart disease
With better cardiothoracic surgical techniques and
technologic advances many pediatric patients with
congenital heart disease survive into adulthood and
present with sequelae of their surgeries and disease.
These include patients with transposition of great
vessels and patients with tetralogy of Fallot. Cardiac
MR imaging can be used for postoperative follow-up
in these patients to assess for occluded shunts (Waterston, Glenn, and so forth); obstructed baffles; and
stenotic homografts. Cardiac MR imaging can also
be used for visualization of previously undetected
disease, such as patent ductus arteriosus, coarctation
of aorta, atrial septal defects, restrictive ventriculoseptal defect, and anomalous pulmonary veins. To
perform cardiac MR imaging for these conditions
one should identify the specific clinical question,
know the anatomy of the relevant pathology and
cardiac surgery, and be aware of the delayed complications typical of the performed surgery. Communication between the referring physician and the
physician performing the cardiac MR imaging examination is essential. In difficult cases one may want
to obtain consultation by teleradiology with a
trained cardiac MR imaging radiologist in an academic center.
General MR imaging protocol for congenital heart
disease
The first sequences obtained are usually blackblood sequences, such as HASTE (double IR
FSE-TSE), or TSE-FSE T1-weighted sequences.
Bright-blood sequences, such a sequential FLASH,
FASTCARD, trueFISP, or FIESTA, are essential
for demonstrating functional pathology and may be
necessary to visualize some intracardiac shunts. Cine
sequences (GRE) should be done, at the very least,
through the area of suspected pathology. Ideally,
depending on the disease, a congenital heart disease
protocol includes some transverse imaging (ie, the
black-blood scout) to assess the great vessels (ie, presence of a duplicated superior vena cava, sidedness of
the arch), and four-chambered long-axis black-blood
and cine sequences. Cine sequences should then be
performed through the aortic and pulmonic valve
planes and through any surgically created shunts
(Fontan, Waterston, Blalock-Taussig, and so forth) to
assess for patency and stenoses. Contrast-enhanced
MR angiography can be used to assess peripheral
pulmonary artery stenoses; bronchial collaterals (pulmonic atresia); or anomalous pulmonary veins.
Contrast-enhanced MR angiography
In the cardiac MR imaging assessment of adult
patients with congenital heart disease, contrast-enhanced MR angiography is useful for evaluation of
the aorta, pulmonary artery stenoses, collaterals, and
shunts (Fig. 9). Contrast-enhanced MR angiography
is a short breathhold three-dimensional GRE sequence with short TR and TE and flip angle. No cardiac gating is needed. It requires a test-bolus injection
or bolus tracking system, such as CareBolus (Siemens Medical Systems, Erlangen, Germany) or
SmartPrep (GE Medical Systems, Milwaukee, Wisconsin), to calculate the circulation time and obtain
images with maximum arterial enhancement. Injection rate is usually 2 mL/second of 0.2 mmol/Kg
Gd-DTPA. A commercially available MR imaging
compatible power injector is required. Images are
usually obtained in a coronal orientation, but can also
be obtained in an oblique-sagittal orientation to assess
the aortic arch. Both precontrast and postcontrast
images are acquired with the precontrast image
serving as a mask for image subtraction. After image
acquisition, postprocessed three-dimensional maximum intensity projection images can be created.
These maximum intensity projection images should
always be evaluated together with source images to
avoid misdiagnoses secondary to maximum intensity
projection induced artifacts.
Newer sequences that allow near real-time assessment of dynamic administration of a gadoliniumbased contrast bolus are now available. These sequences, although by necessity of lower resolution
than non real-time sequences, are helpful in the
assessment of shunts and fistulas [13].
Transposition of great arteries
In D-loop transposition (Fig. 10), the anatomic
relationship of great arteries is reversed. The aortic
valve arises anterior to the pulmonic valve. Aortic
valve and aorta arise from the right ventricle, which is
usually hypertrophied. The pulmonary valve and
pulmonary artery arise from the left ventricle. In
L-loop transposition, the aorta is left sided and arises
from the right ventricle, which may at times be
rudimentary. The pulmonary artery arises posteriorly
and to the right of the aorta from the left ventricle. If
the two ventricles are well-developed and there is no
interventricular communication this entity is referred
to as congenitally corrected transposition of the
great vessels. At times, however, there is a large
ventriculoseptal defect with a rudimentary right ven-
505
Fig. 9. A 20-year-old female patient with hypertension, unresponsive to medication. Lateral (A) and coronal (B) views of
contrast-enhanced three-dimensional MR angiography clearly show coarctation of descending aorta (arrow) and extensive
collateral vessels.
tricle underneath the aorta. The great arteries are

transposed and the two ventricles are inverted. Repair
of D-transposition of great vessels is currently performed using an arterial switch procedure (Jatene).
Previously, a Mustard or Senning baffle operation was
used. In the Mustard operation, the intra-atrial septum
is removed and bovine pericardium creates anterior
(systemic) and posterior (pulmonary venous) baffles
(Fig. 11). Most often the reason for performing an
MR imaging examination in these patients who
have had a baffle is evaluation for baffle patency,
including evaluation for possible stenoses that may
develop at the superior vena cava as it enters the
superior limb of systemic baffle, or at the pulmonary veins.
Fig. 10. D-loop transposition of the great arteries. Axial

dark-blood HASTE. Note the aorta arising anterior and to
the left of the main pulmonary artery. Ao, aorta; PA, pulmonary artery.
Tetralogy of Fallot
The classic components of tetralogy of Fallot are a
large ventricular septal defect, right ventricular outflow tract obstruction, right ventricular hypertrophy,
and overriding aorta. Complete repair of tetralogy of
506
origin. The ductus usually closes shortly after birth. It

can remain patent, however, and occasionally can
present in the adult patient. Cine GRE MR imaging
sequences can evaluate for the presence of patent
ductus arteriosus, or one can use a contrast-enhanced
GRE technique. As illustrated in Fig. 13, the performance of nontraditional off-axis or orthogonal-plane
images may help further to evaluate pathology.
Fig. 11. Transaxial TrueFISP cine image shows a Mustard

correction of D-loop transposition of the great arteries. The
superior and inferior vena cavae supply blood to the anterior
baffle, which directs flow to the left ventricle and pulmonary
arteries. The pulmonary veins supply blood to the posterior
baffle, which directs flow to right ventricle and aorta. Note
that the right ventricle, now the principle pumping chamber,
has become large and thick walled. A, anterior baffle; LV,
left ventricle; P, posterior baffle; RV, right ventricle.
Fallot is achieved with ventriculoseptal defect closure

and infundibulectomy. Before complete repair, some
patients undergo a palliative shunt to improve pulmonary blood flow. Shunts that have commonly been
performed include the Blalock-Taussig, Waterston,
Potts, and Glenn anastomosis. The Blalock-Taussig
shunt connects the subclavian artery to the pulmonary
artery. For evaluation of this shunt with MR imaging
an oblique transverse plane is obtained. The Waterston shunt connects the ascending aorta to right
pulmonary artery. The Potts shunt connects the descending aorta and left pulmonary artery. The Waterston and Potts shunts are best evaluated with MR
imaging in transverse plane. The Glenn shunt connects the superior vena cava to the right pulmonary
artery. Coronal images in the plane of the superior
vena cava are useful for MR imaging evaluation of
this shunt. MR imaging examination in patients with
shunts is often performed to assess for shunt stenosis
or occlusion (Fig. 12). MR imaging also can assess
for the presence of a stenosis at homograft anastomosis or valve.
Patent ductus arteriosus
The ductus arteriosus is a normal tubular structure
that connects the underside of the descending aorta
just distal to the origin of the left subclavian artery to
the main or left pulmonary artery just beyond its
Ventricular septal defect

The anatomic location of intracardiac shunts, such
as ventriculoseptal defect or atrial septal defect, can
be demonstrated definitely by MR imaging. In assessing for a small or restrictive ventriculoseptal defect,
atrial septal defect, or patent foramen ovale GRE cine
sequences are vital to visualize the jet caused by
turbulent flow. The shunt may be missed if only
black-blood anatomic imaging is obtained.
Anomalous pulmonary veins
Because of its multiplanar capability, MR imaging
is highly accurate in the diagnosis of partial or total
anomalous pulmonary venous connection and several
other anomalies of the venous system. Anomalous
right upper lobe pulmonary vein (Fig. 14) usually
drains into superior vena cava and often is associated
with a sinus venosus atrial septal defect. An anomalous left upper lobe pulmonary vein may look like a
duplicated superior vena cava, but can be differenti-
Fig. 12. Axial cine GRE image shows a stenotic Waterston

shunt in a patient with pulmonic atresia and hypoplastic
right ventricle. The Waterson shunt connects the ascending
aorta to the right pulmonary artery. Arrow points to a jet
in right pulmonary artery.
507
Fig. 13. (A) Transaxial true FISP image shows an abnormal structure adjacent (arrow) to the aorta and superior pulmonary trunk.
(B) True FISP sagittal image obtained in plane through the long axis of the structure seen on image A provides greater
characterization and demonstrates that the structure is a patent ductus. Turbulent flow through the ductus produces a jet into the
pulmonary trunk (arrow).
Fig. 14. Anomalous right upper lobe pulmonary vein. Young woman with Turners syndrome who had an enlarged right atrium
seen on an echocardiogram. The etiology of the enlarged right atrium could not be determined. An oblique cine image
(A) showed a dilated superior vena cava with a small jet (arrow). Additional imaging in the plane of the jet (B) showed an
anomalous right upper lobe pulmonary vein (arrow).
508
ated from superior vena cava duplication by following the vessel back to its origin.
Arrhythmogenic right ventricular dysplasia
One of the more frequent and important indications for cardiac MR imaging is the evaluation of patients with potential diagnosis of arrhythmogenic right
ventricular dysplasia. This condition is a primary disorder of the right ventricle with partial or total thinning and replacement of muscle by adipose or fibrous
tissue and enlargement of the right chambers of the
heart. Patients have ventricular arrhythmias and left
bundle branch block on ECG. The disease may lead
to sudden death. Right ventricular dysplasia is familial in 30% of cases. Inheritance pattern is possibly
autosomal-dominant with variable expression and
penetrance [14 16].
Right ventricular angiography and echocardiography cannot visualize pathologic structural changes
of right ventricular dysplasia in the myocardium.
Even with endomyocardial biopsy the diagnosis can
be difficult, because the disease rarely involves the
septum, which is the typical sampling site. Patients
are commonly referred for cardiac MR imaging
[17,18].
MR imaging findings in right ventricular dysplasia
MR imaging diagnosis is based on the identification of specific anatomic and functional abnormalities
of the right ventricle, which include the following
(Figs. 15 and 16): thinning of the right ventricular
free wall; increased myocardial signal intensity from

fatty replacement; decreased systolic wall thickening
or motion (right ventricular akinesia or dyskinesias)
causing focal bulging at the site of myocardial fibrosis; diminished ejection fraction; and impaired ventricular filling in diastole. The right ventricle and
atrium can be normal in size or dilated.
MR imaging sequences
MR imaging evaluation of right ventricular dysplasia is achieved by using black-blood breathhold
sequences along with bright-blood cine imaging.
Usually one plane is obtained in either the long-axis
or transverse image orientation, contiguous 5-mm
slices with no gap. Sagittal or short-axis images may
also be useful. Cine sequences are very important to
assess for areas of right ventricular dysfunction (akinesia, dyskinesias, and focal bulge). Breathhold FSE
(FSE-TSE T1) also can be performed with and without fat saturation through areas of suspicion. These
images often provide sharper images than available
with dark-blood HASTE, but have the disadvantage in
that usually only a single slice can be obtained during
a breathhold period.
Constrictive pericarditis versus restrictive
cardiomyopathy
Because both entities have similar clinical signs
and symptoms, MR imaging can be particularly
useful in differentiating between constrictive pericarditis and restrictive cardiomyopathy. MR imaging is
Fig. 15. Biopsy-proved case of arrhythmogenic right ventricular dysplasia. (A) axial TSE T1-weighted image shows fatty
infiltration of the myocardium involving the pulmonary outflow tract (arrow). (B) Fat-saturated axial TSE T1-weighted image
shows signal dropout of this region (arrow), caused by fatty infiltration.
Fig. 16. Axial bright-blood cine GRE image in another

patient with right ventricular dysplasia shows focal bulge
(arrow) of the right ventricular wall.
very useful in making this distinction because the

pericardium has normal thickness in restrictive cardiomyopathy [19,20]. Constrictive pericarditis results
from progressive pericardial fibrosis and calcification, leading to restriction of cardiac ventricles during
diastole. Constriction may follow any pericardial
injury that causes an inflammatory response, such
as infectious pericarditis, connective tissue disease,
neoplasm, renal failure, cardiac surgery, and radiation
therapy. The normal pericardium is very thin (1 to
2 mm). A thickness of 4 mm or more indicates
pericardial thickening and in proper clinical setting
is the finding that is diagnostic of constrictive pericarditis (Fig. 17). Other associated findings are
markedly dilated inferior vena cava, hepatic veins,
and right atrium. Right ventricle has normal or
reduced volume. Restrictive cardiomyopathy is uncommon and results from infiltrative conditions leading to myocardial stiffness and restriction. Causes
include both infiltrative (amyloid, sarcoid) (Fig. 18)
and noninfiltrative (idiopathic, scleroderma) processes; storage diseases; and carcinoid and endomyocardial fibrosis. Besides normal pericardial thickness, the
myocardium is thickened. Like constrictive pericarditis, patients with restrictive cardiomyopathy may
also demonstrate enlarged atria and dilatation of the
inferior vena cava and hepatic veins [21].
The wall thickness of either or both ventricles is
usually increased in the restrictive cardiomyopathy
509
Fig. 17. Constrictive pericarditis. Axial dark-blood HASTE

image shows a thickened pericardium ( > 4 mm); normal
myocardial thickness; large right and left atria; and relatively
small ventricular size. MR imaging may also show a dilated
inferior vena cava or paradoxical septal motion because of
increased right-sided pressure.
associated with amyloidosis [22]. Restrictive cardiomyopathy is frequently complicated by mitral or

tricuspid regurgitation. This can be demonstrated
and quantified using cine GRE [23]. Stasis of blood
in atria can cause high signal intensity within the
atrium on older spin echo images.
Constrictive pericarditis and restrictive cardiomyopathy both have similar clinical signs and symptoms: right-sided heart failure; peripheral edema;
distended neck veins; and Kussmauls sign (venous
pressure fails to drop with inspiration). MR imaging
Fig. 18. Patient with sarcoid. Axial dark-blood HASTE

shows normal thickness of pericardium; markedly thickened
myocardium; small ventricular volume; and, like restrictive
pericarditis, large right and left atria.
510
can help cardiologists to differentiate between these

two conditions and may be especially useful because
constrictive pericarditis can be treated surgically by
stripping the pericardium.
Black-blood imaging can be performed in both
four-chamber long-axis and short-axis planes. As with
imaging for right ventricular dysplasia, breathhold
TSE T1-weighted images may prove more useful than
HASTE or double IR TSE-FSE images, which tend
to be a bit grainy. Bright-blood cine sequences are
also useful, especially in the assessment of myocardial
thickness if restrictive cardiomyopathy is suspected.
These can be performed in the short axis for both
qualitative analysis of myocardial thickness, or quantitative analysis of myocardial thickness, mass, and
left ventricular stroke volume using commercially
available analysis packages.
It is important to know that calcification of the
pericardium in constrictive pericarditis is dark on MR
images. If needed, noncontrast CT can help visualize
pericardial calcification in patients without demonstrable pericardial thickening, but with a high clinical
suspicion of constrictive pericarditis.
Cardiac tumors and metastatic disease
Primary cardiac tumors are rare and approximately 80% are benign. Secondary tumors involving
the heart are 40 to 50 times more frequent than
primary tumors [24]. In general most metastasis and
malignant tumors are broad-based or invade the

myocardium. Most benign tumors are intraluminal
and are attached by a narrow stalk. Most tumors
enhance with gadolinium. This helps to differentiate
them from thrombus, which does not enhance.
Primary tumors of the heart include myxoma (the
most frequent benign cardiac tumor, usually within
the left atrium) (Fig. 19); lipoma (usually right atrium,
fat saturation helpful); angiosarcoma (most commonly malignant, arising from the right atrium);
rhabdomyoma (frequent tumor in children); fibroma
(low signal on T2); and hemangioma light bulb appearance on T2-weighted images. Secondary tumors
of the heart include hematogenous metastatic disease
to myocardium and pericardium, but most frequently
metastatic disease as an extension from tumors of
the adjacent lung or mediastinal structures. Extension
of tumors of the upper abdomen can also occur
through the inferior vena cava into the right atrium.
The most common mass of the heart in general is a
thrombus, which most frequently involves the left
atrium or ventricle.
It is important to note that some normal cardiac
anatomic structures may be confused with thrombus
or mass on both MR imaging and echocardiography,
and it is important to recognize these normal structures as such. The crista terminalis and associated
Chiari network, a nodular filamentous structure that
runs along the posterior aspect of the right atrium, may
be mistaken as thrombus. Lipomatous hypertrophy of
the interatrial septum is echogenic on echocardiography, and also may be mistaken as a mass, but clearly
Fig. 19. Right atrial myxoma. (A) Axial dark-blood HASTE shows a bilobed mass (arrow) straddling the tricuspid valve. Note
the relatively bright signal of the mass on this T2-weighted sequence. Images are acquired in diastole and do not demonstrate the
location of the mass throughout the cycle. (B) Axial bright-blood cine GRE image obtained in systole shows that the mass arises
from the right atrium with the point of tumor attachment at the intra-atrial septum (arrow).
can be identified as fat by its signal characteristics

(bright on T1 and T2, signal suppressed with spectral
fat saturation) on MR imaging.
Dark-blood HASTE (double IR TSE-FSE) in one
or more planes is useful in evaluation of tumors. This
sequence is suggested because of its T2-weighting,
because most tumors have high signal intensity on
T2-weighted images. GRE cine sequences should
also be performed because they are helpful in the
assessment of the tumor attachment point (narrow
versus broad-based). Breathhold TSE T1 or T1weighted GRE sequences with an IR pulse (similar
to that used for myocardial viability assessment)
pregadolinium and postgadolinium administration
can also be used to determine extent of vascularity
and enhancement [24].
Clinical coronary MR angiography
Although coronary MR angiography for atherosclerosis assessment is currently a research examination, coronary MR angiography can be applied for
certain clinical reasons. Current clinical applications
of coronary MR angiography include assessment of
anomalous coronary arteries, coronary artery aneurysm, and assessment of bypass graft patency. Some
congenital anomalous coronary artery arrangements
are associated with sudden death [25]. Anomalous
coronary arteries associated with sudden death include a right coronary artery or left anterior descending coronary artery traveling between the aorta and
pulmonic outflow tract. Some of these anomalies are
also difficult to evaluate with conventional X-ray
coronary angiography. Because of its multiplanar
imaging capabilities MR imaging is useful and can
help as a problem-solving tool to evaluate the exact
pathway of an anomalous coronary artery in ways
conventional X-ray angiography cannot.
Techniques
Because coronary arteries are small tortuous structures subject to continuous respiratory motion and
cardiac contraction, it is difficult in general to image
the coronary arteries with MR imaging. With the
development of commercially available new ultrafast
imaging techniques, however, excellent quality MR
angiography of the coronary arteries has become
feasible to perform clinically. Three-dimensional
methods are usually the most useful and now can
be performed with either breathhold or respiratory
gating. Techniques most often used are standard GRE
techniques, with or without contrast enhancement;
511
however, newer TrueFISP-type sequences provide

excellent coronary artery signal. Black-blood methods have also been used. Two-dimensional breathhold GRE cines can be useful for assessing bypass
graft patency.
Valvular disease
Quantification of blood flow through heart valves
is of clinical interest in the assessment of the severity
of valvular heart disease. Although valvular stenosis
may be adequately evaluated by measuring transvalvular pressure gradients using Doppler cardiac echo
or cardiac catheterization, traditional methods fail to
provide consistently reliable and accurate quantification of valvular regurgitation [26]. Cine MR imaging has been found to be an effective technique
for evaluating ventricular and valvular function in
certain valvular heart diseases [1,23,27 31].
Cardiac MR imaging techniques can demonstrate
the presence and quantify the severity of valvular
heart disease. MR imaging examination of valvular
dysfunction includes direct demonstration of the jet
of valvular stenosis or regurgitation and demonstration of chamber dilation or hypertrophy. Cine MR
imaging displays signal void in areas of turbulent
flow related to either valvular disease, such as stenosis or insufficiency (Fig. 20). Size of signal loss is
dependent on degree of turbulent flow and on chosen
echo time (TE). Velocity-encoded cine MR imaging
can be used for measurement of peak velocities
through the area of stenosis. Care must be taken to
use a sequence with a velocity-encoded cine above
the estimated peak velocity to avoid inaccuracies
caused by aliasing. Using a modified Bernoulli equation (eP = 4V2), the pressure gradient (eP) across the
valve or stenotic segment of vessel can be estimated
when systolic peak velocity (V) is known. A pressure
gradient more than 25 mm Hg is hemodynamically
significant. The regurgitant fraction in aortic valve
insufficiency can be determined by calculating the
difference of the right and left ventricular stroke
volumes. This is only accurate if no shunt or other
valve disease is present [31].
The expanding role of cardiac MR imaging

The role of cardiac MR imaging in the evaluation
of heart disease has expanded from the traditional
role of anatomic characterization toward functional
evaluation. Some of the new clinical applications of
cardiac MR imaging include myocardial function
studies, myocardial viability, and coronary MR angi-
512
Fig. 20. (A) Sagittal breathhold single-slice TSE T1-weighted image in a young woman with Takayasus arteritis. Note the aortic
wall thickening (arrow). (B) Coronal breathhold cine GRE in the same patient shows a jet of aortic insufficiency (arrow) through
the aortic valve toward the left ventricle. The aortic insufficiency is caused by poor apposition of the aortic valve leaflets. Note
the dilatation of the sinuses at the aortic root.
ography for evaluation of proximal stenosis of coronary arteries.

Left ventricular function
Cine GRE sequences permit assessment of left
ventricular function and can be performed during rest
and pharmacologic stress. To perform an MR imaging examination for assessment of left ventricular
function short-axis cine GRE breathhold sequences
are performed contiguously from the base of the
left ventricle through the left ventricular apex. Retrogated sequences are ideal because they allow for
imaging throughout the complete R-R interval. Using
commercially available software, one can select the
images obtained at end-systole or end-diastole; segment the blood pool from the myocardium; and
calculate ejection fraction, left ventricular volume,
and myocardial mass. Newer completely gradientrefocused bright-blood sequences (TrueFISP, balanced fast field echo, FIESTA) are ideal for use with
these packages because of the increased contrast
between blood pool and myocardium in comparison
with standard GRE sequences.
Fig. 21. Saturation-tagged short-axis image of the heart. The

dark lines are saturation bands forming a grid (arrow),
which is deformed during normal myocardial contraction.
Left ventricular wall motion abnormalities

Left ventricular wall motion abnormalities can be
assessed qualitatively, but also quantitatively using
saturation tagged sequences and mathematical models. Qualitative evaluation of wall motion abnormalities is usually not required clinically. Fig. 21 shows
the typical tagged cardiac sequence in short-axis view.
The dark lines are tagged lines (saturation bands)
forming a diamond, which deforms during ventricular
muscle contraction.
Myocardial perfusion and viability

Myocardial perfusion studies are performed using
first-pass myocardial TurboFLASH-type sequences
capable of rapidly establishing T1 contrast for multiple slices with high temporal resolution. Rapid
administration of intravenous gadolinium, to provide
a tight bolus, is administered during pharmacologic
stress (intravenous infusion of adenosine). Depending on the R-R interval, three to five slice positions
can be obtained over multiple phases, demonstrating
low signal areas of underperfusion in the myocardium. These low signal areas correspond to regions
of ischemia or infarct. Patients must be hemodynamically monitored during adenosine infusion, with
a lead II rhythm strip monitored for complete heart-
513
block and pulse oximetry and visual examination for

bronchospasm. The adenosine and gadolinium-based
MR imaging contrast agent, although administered
simultaneously, should be administered through separate intravenous cannulae in separate arms, to avoid a
bolus push of the adenosine from the rapidly administered intravenous contrast agent.
Single slice IR TurboFLASH, TrueFISP, or newer
three-dimensional IR-prepared sequences can then be
used in a delayed fashion (usually 10 15 minutes
after contrast injection) to demonstrate regions of
delayed contrast washout corresponding to infarcted
tissue. Fig. 22 shows an area of infarcted myocardium
that enhances on delayed contrast-enhanced T1 image
(segmented Turbo FLASH technique) [2,32]. With
these sequences, selection of the IR pulse is key to
suppress signal from (null) the normal myocardium.
Although usually between 200 and 300 milliseconds,
the optimal IR pulse is dependent on the dose of
contrast administered and the length of delay after
contrast injection.
Summary
Because of the enormous economic and social
impact of cardiovascular disease in the United States
there is a need for improved noninvasive diagnosis.
Cardiac MR imaging is a versatile, comprehensive
technique for assessing cardiac morphology and
function. With an understanding of cardiac anatomy
and physiology and MR imaging physical principles,
cardiac MR imaging can be performed and can play
an important role in patient management.
References
Fig. 22. Subendocardial lateral wall myocardial infarction.

Delayed imaging in the short axis is obtained using a
contrast-enhanced T1-segmented TurboFLASH sequence.
Arrows point to the nonviable myocardium.
[1] Wagner S, Auffermann W, Buser P, Semelka RC,

Higgins CB. Functional description of the left ventricle
in patients with volume overload, pressure overload,
and myocardial disease using cine nuclear magnetic
resonance imaging (NMRI). Am J Card Imaging
1991;5:87 97.
[2] Kim RJ, Fieno DS, Parrish TB, Harris K. Relationship
of MRI delayed contrast enhancement to irreversible
injury, infarct age, and contractile function. Circulation
1999;100:1992 2002.
[3] Wilke N, Jerosch-Herold M, Wang Y, Huang Y, Christensen BV, Stillman AE, et al. Myocardial perfusion
reserve: assessment with multisection, quantitative,
first-pass MR imaging. Radiology 1997;204:373 84.
[4] Atkinson D, Teresi L. Magnetic resonance angiography (review article). Magn Reson Q 1995;10:149 72.
[5] Finn JP, Edelman RR. Black-blood and segmented
514
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]

k-space magnetic resonance angiography. Magn Reson
Imaging Clin N Am 1993;1:349 57.
Woodard PK, Zheng J. Congenital heart disease, Unit
A10. In: Haacke EM, Lin W, editors. Current protocols
in magnetic resonance imaging. New York: John Wiley
and Sons; 2001. p. A10.1.1 10.1.19.
Lee VS. Pericardial disease, Unit A11A. In: Haacke
EM, Lin W, editors. Current protocols in magnetic
resonance imaging. New York: John Wiley and Sons;
2001. p. A11.1.1 11.1.14.
Boxt LM. How to perform cardiac MR imaging. In:
Boxt LM, editor. Cardiac MR imaging. Magn Reson
Imaging Clin N Am 1996;4:191 216.
Pereles FS, Kapoor V, Carr JC, Simonetti OP, Krupinski
EA, Baskaran V, et al. Usefulness of segmented trueFISP cardiac pulse sequence in evaluation of congenital
and acquired adult cardiac abnormalities. AJR Am J
Roentgenol 2001;177:1155 60.
Rademakers FE, Bogaert J. Cardiac anatomy. In:
Bogaert J, Duerinckx AJ, Rademakers FE, editors.
Magnetic resonance imaging of the heart and great
vessels. Berlin: Springer; 1999. p. 29 38.
Dinsmore RE, Wismer GL, Levine RA, Okada RD,
Brady JJ. MRI of the heart: positioning and gradient
angle selection for optimal imaging plane. AJR Am J
Roentgenol 1984;143:1135 42.
Buser PT, Auffermann W, Holt WW, Wagner S,
Kircher B, Wolfe C, et al. Noninvasive evaluation of
global left ventricular function with cine NMR. J Am
Coll Cardiol 1989;13:1294 300.
Finn JP, Baskaran V, Carr J, McCarthy RM, Pereles
FD, Kroeker R, et al. Thorax: low dose contrastenhanced three-dimensional MR angiography with
subsecond temporal resolutioninitial results. Radiology 2002;224:896 904.
Basso C, Thiene G, Corrado D, Angelini A, Nava A,
Valente M. Arrhythmogenic right ventricular cardiomyopathy: dysplasia, dystrophy, or myocarditis? Circulation 1996;94:983 91.
Corrado D, Basso C, Thiene G, et al. Spectrum
of clinicopathologic manifestations of arrhythmogenic
right ventricular cardiomyopathy/dysplasia: a multicenter study. J Am Coll Cardiol 1997;30:1512 20.
Fontaine G, Fontaliran F, Frank R. Arrhythmogenic
right ventricular cardiomyopathies: clinical forms
and main differential diagnosis. Circulation 1998;97:
1532 5.
Boxerman JL, Mosher TJ, McVeigh ER, Atalar E,
Lima JA, Bluemke DA. Advanced MR imaging techniques for evaluation of the heart and great vessels.
Radiographics 1998;18:543 64.
White RD, Trohman RG, Flamm SD, et al. Right ven-
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
tricular arrhythmia in the absence of arrhythmogenic

dysplasia: MR imaging of myocardial abnormalities.
Radiology 1998;207:743 51.
Soulen RL, Stark DD, Higgins CB. Magnetic resonance imaging of constrictive pericardial heart disease.
Am J Cardiol 1985;55:480 4.
Masui T, Finck S, Higgins CB. Constrictive pericarditis and restrictive cardiomyopathy: evaluation with
MR imaging. Radiology 1992;182:369 73.
Sechtem U, Higgins CB, Sommerhoff BA, Lipton MJ,
Huycke EC. Magnetic resonance imaging of restrictive
cardiomyopathy. Am J Cardiol 1987;59:480 2.
Borer JS, Henry WL, Epstein SE. Echocardiographic
findings in patients with systemic infiltrative diseases
of the heart. Am J Cardiol 1977;39:184 8.
Wagner S, Affermann W, Buser P, Lim TH, Kircher B,
Pflugfelder P, et al. Diagnostic accuracy and estimation
of the severity of valvular regurgitation from the signal
void in cine MR. Am Heart J 1989;118:760 7.
Higgins CB. Acquired heart disease. In: Higgins CB,
Hricak H, Helms CA, editors. Magnetic resonance
imaging of the body. 3rd edition. Philadelphia: Lippincott-Raven; 1997. p. 409 60.
Felmeden D, Singh SP, Lip GY. Anomalous coronary
arteries of aortic origin. Int J Clin Pract 2000;54:390 4.
Carroll JD. Cardiac catheterization and other imaging
modalities in the evaluation of valvular heart disease.
Curr Opin Cardiol 1993;8:211 5.
Sechtem U, Pflugfelder PW, White RD, Gould RG,
Holt W, Lipton MJ, et al. Cine MRI: potential for
the evaluation of cardiovascular function. AJR Am J
Roentgenol 1987;148:239 46.
Dulce MC, Mostbeck G, OSullivan MM, Cheitlin MD,
Caputo GR, Higgins CB. Severity of aortic regurgitation: interstudy reproducibility of measurements with
velocity-encoded cine MR imaging. Radiology 1992;
185:235 40.
Fujita N, Chazouilleres AF, Hartiala JJ, OSullivan M,
Heidenreich P, Kaplan JD, et al. Quantification of mitral
regurgitation by velocity-encoded cine magnetic resonance imaging. J Am Coll Cardiol 1994;23:951 8.
Globits S, Higgins CB. Assessment of valvular heart
disease by magnetic resonance imaging. Am Heart J
1995;129:369 81.
Didier D, Ratib O, Lerch R, Friedli B. Detection and
quantification of valvular heart disease with dynamic
cardiac MR imaging. Radiographics 2000;20:1279 99.
Kim RJ, Chen EL, Lima JA, Judd RM. Myocardial
Gd-DTPA kinetics determine MRI contrast enhancement and reflect the extent and severity of myocardial
injury after acute reperfused infarction. Circulation
1996;94:3318 26.
Radiol Clin N Am 42 (2004) 515 541
Plain film / MR imaging correlation in heart disease

Andre J. Duerinckx, MD, PhDa,b,*
a
Radiology Service, Veterans Affairs North Texas Healthcare System, 4500 South Lancaster Road, Dallas, TX 75126, USA
b
University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA
The plain film of the chest often is used as a first

screening test for a wide range of diseases in the
thorax. Cardiovascular disease and specifically heart
disease are often first diagnosed or suggested based
on findings on the plain film of the chest [1 3].
Although the severity of some cardiovascular diseases is readily reflected on the chest film, other
important diseases are barely or not visible on plain
films. When a plain film of the chest suggests cardiacor vascular anomalies a follow-up cross-sectional
study is often performed. Cross-sectional imaging
of the chest is done with echocardiography, CT, or
MR imaging. The newer CT and MR imaging technologies allow the acquisition of high-resolution and
high-quality angiograms of the thoracic vessels using
CT angiography or MR angiography with or without
three-dimensional volume rendering [4 11]. CT and
MR imaging also allow almost real-time high-resolution imaging of the heart and provide anatomic and
functional information [12 16]. This article reviews
common cardiovascular pathologies that can be noted
first on plain film when previously unsuspected. Next
illustrated is how cross-sectional imaging can provide
the follow-up information needed to make a diagnosis. Even though the article mostly shows examples
with MR imaging cross correlation, it should be kept
in mind that in 2004 cardiac CT and CT angiography
are often equally acceptable for the follow-up of
cardiovascular anomalies in the chest. First reviewed
are the normal cardiac structures and contours as seen
on the plain film of the chest. Then reviewed are
specific types of pathologies as seen in older adults;
* VA North Texas Healthcare System, Radiology Service (W-114), 4500 South Lancaster Road, Dallas, TX 75126.
E-mail address: andre.duerinckx@med.va.gov
patients with lung cancer invading the heart, pericardium, or large vessels; and postsurgical and posttraumatic findings. Also provided is a review of non
cardiac-related areas of plain film and cross-sectional
imaging correlation. Understanding this correlation
for thoracic imaging is one of the ideas and principles
outlined in the proposal for a curriculum in cardiothoracic radiology for medical students in the year
2000 by Kazerooni et al [17] and Collins et al [18]. It
is hoped that the reader gains a better understanding
and appreciation for the great value of cross-sectional
imaging and the power of the plain film in helping
detect and recognize thoracic pathology.
Normal cardiac structures as seen on the plain

film
The normal cardiac structures and their appearance on plain films have been well described in the
literature [1,3,19,20]. These structures include the
right atrium, the left atrium, the right ventricle, and
the left ventricle. The expected appearance of the
normal heart on a plain film is well known, and many
detailed descriptions can be found in most radiology
textbooks dealing with thoracic and cardiac imaging
[2]. This normal cardiac silhouette is discussed relying on a review by Baron [3].
The frontal (posteroanterior or anteroposterior)
view is invaluable for cardiac evaluation. It provides
information about cardiac situs, great vessels, pulmonary vascularity, and the heart (Fig. 1). The overall
cardiac size and certain types of chamber enlargement
may be determined. The right heart border is roughly
divided into two equal parts: an upper, straight margin formed by the wall of the superior vena cava
(SVC) and a lower, convexly curved segment repre-
doi:10.1016/j.rcl.2004.03.001
516
A.J. Duerinckx / Radiol Clin N Am 42 (2004) 515541
Fig. 1. A 59-year-old man with question of dilated aorta. Correlation between frontal radiograph of the chest (A) and the frontal
view of a maximum intensity projection reconstruction of a three-dimensional contrast-enhanced MR angiogram of the thoracic
vessels (B). (C) An oblique sagittal (candy cane) view of the MR angiogram of the thoracic aorta is also shown. There is
excellent correlation between the appearance of the thoracic aorta and central pulmonary vessels on both the plain film and the
MR angiogram. The ascending aorta on MR image measured 3.1 cm, which is within normal limits.
senting the lateral wall of the right atrium. The left

cardiac contour is composed of four segments. The
aortic knob creates the upper most bulge. This is not
an anatomic structure but simply represents the distalmost portion of the aortic arch as it curves downward to become the descending aorta. The next bulge
represents the main pulmonary artery and a small part
of the outflow tract of the right ventricle. Immediately
beneath this is a small flat, or slightly concave, segment where the left atrial appendix reaches the left
cardiac border. The region of the main pulmonary

segment is above the left bronchus; the region of the
left atrial appendage is immediately adjacent to and
below the left bronchus. The remainder of the left
cardiac silhouette is formed by the broad curve of the
lateral wall of the left ventricle. The right ventricle
and ascending aorta do not normally contribute to the
cardiac borders on the frontal projection. The lateral
view is particularly helpful in assessing the right ventricle and left atrium.
Abnormal cardiac structures seen on plain film

A systematic approach to the chest radiograph to
evaluate cardiovascular disease involves evaluation
of thoracic musculoskeletal structures, pulmonary
vascularity, overall heart size, specific chamber enlargement, and the great arteries. It is important to
recognize signs of pulmonary venous hypertension,
radiographic features of specific cardiac lesions,
features of big heart disease, and abnormal cardiac contours [21]. These basics are well covered
in several classic textbooks and elsewhere in this
issue [1,2].
Abnormalities of the superior vena cava segment

The dilatation of the ascending aorta can create a
diffuse outward bowing of the SVC segment. This
can be seen with a variety of diseases of the thoracic
aorta. With degenerative disease, the aorta can become tortuous and ectatic (Fig. 2). Abnormalities in
cardiac function, such as an increased cardiac stroke
volume caused by aortic insufficiency, can also create
dilatation of the aorta. The appearance of poststenotic
dilatation secondary to aortic valve stenosis is typically limited to the mid third of the SVC segment.
Rupture of the right atrium SVC junction can be
seen in cases of blunt thoracic trauma [22]. Tumors,
such as asymptomatic lipoma of the SVC, can
present as mediastinal widening on a chest radiograph [23].
Enlargement of the right atrial segment

Enlargement of the heart to the right can be caused
by dilatation of either atrium. Most common is
enlargement of the right atrium, which causes the
right contour to become increasingly convex and to
bow outward in a single curve from a SVC segment
to the diaphragm. In some cases, however, the left
atrium can become so enlarged as to extend behind
the right atrium and reach the right cardiac border.
This then creates the typical bilobed configuration
with the left atrium forming the upper bulge and the
supradiaphragmatic component representing the right
atrium. There are limitations, however, as to what the
plain film can reveal [24].
Other causes of enlargement of the right heart
contour or abnormalities along the contour are absence of the right pericardium, pericardial cyst, pericardial fat pad, pericardial tumor, and mediastinal
tumor [21].
517
Enlargement, calcification, widening, or blurring of

the aortic knob segment
Enlargement of the aortic knob is mostly caused
by diseases of the aorta, such as aneurysm or degenerative disease (Fig. 3). The significance of calcification of the aortic knob has been extensively
discussed [25 30]. In an adult Greek population
(1027 patients) the gravity of aortic arch calcification
has been assessed easily on routine chest radiograph
and was positively correlated with coronary artery
disease and important cardiovascular risk factors
(age, diabetes mellitus, hypertension, and dyslipidemia) [30]. Widening of the aortic knob has been
described in aortic dissection [31]. Blurring of the
aortic knob has been described in posttraumatic
rupture of the aorta [32,33]. Blurring of the left
hilum also has been described after aortic rupture
[34]. Chest radiographs obtained at admission of
patients with hemorrhage from ruptured thoracic
aorta aneurysms, aortic dissections, or penetrating
aortic ulcers can show obscuration or convexity of
the aorticopulmonary window; enlarged aortic knob
width; enlarged thoracic aorta size; or an enlarged,
obscured, or irregular aortic margin [35]. These
and other related findings are described in greater
detail later.
Enlargement of pulmonary artery segments
The prominence of the normal pulmonary artery
seen in younger individuals decreases with age and
presents only a minimal convexity in individuals over
the age of 35 to 40 years. Abnormal dilatation of the
pulmonary artery, regardless of age, is usually indicative of pulmonary arterial hypertension. Pulmonary
hypertension is caused by two basic mechanisms:
increased blood flow through the pulmonary circulation, as in left-to-right shunts; or increased resistance
to flow with dilatation of the upstream vessels. The
two types of pulmonary hypertension can often be
differentiated from the chest film. If the cause of
pulmonary hypertension is increased flow, then the
peripheral pulmonary arteries are dilated, along with
the main pulmonary artery and hilar vessels. If the
cause of pulmonary hypertension is resistive hypertension, however, then the peripheral vessels are
usually constricted. Causes of resistive hypertension include vascular occlusion secondary to repeated
emboli [36,37]; compression of the capillary bed
because of lung disease, such as chronic obstructive
pulmonary disease; idiopathic thickening of the arterial walls in primary pulmonary hypertension; and so
forth. Although this can often be well appreciated on
518
Fig. 2. A 62-year-old man with chest pain. Correlation between frontal radiograph of the chest (A) and the frontal view of a
maximum intensity projection reconstruction of a three-dimensional contrast-enhanced MR angiogram of the thoracic vessels
(B). (C) An oblique sagittal (candy cane) view of the MR angiogram of the thoracic aorta is also shown. When compared with the
aorta in Fig. 1, the increased tortuosity is well seen on the plain film. Incidental note is made of a bovine aortic arch. The
ascending aorta measured 3 cm, within normal limits.
plain films it is even better appreciated on crosssectional images [38 50].

There are other causes of pulmonary artery dilatation, such as pulmonary valve stenosis, which is a
common congenital cardiac lesion often first detected
in adult life. The poststenotic dilatation caused by a
high velocity systolic jet, directed posteriorly and
to the left, causes a combination of a dilated main
pulmonary artery with abnormal prominence of

the left pulmonary artery. This is almost pathognomonic for pulmonic valve stenosis. Unfortunately,
the prominence of the left pulmonary artery is not
always present.
Pulmonary artery aneurysms can also present as
masses on the plain film [51 55]. Janssens et al [53]
have reported a case of proximal pulmonary artery
519
Fig. 3. A 77-year-old man with anemia and weight loss. (A, B) Frontal and lateral chest radiographs show mediastinal widening
with prominence of the aortic knob, and with a supra-aortic mass best seen on the lateral view. (C) Oblique sagittal (candy cane)
view from a MR angiogram shows a saccular aneurysm arising superiorly from the distal aortic arch, and corresponding to the
shadow seen on the lateral chest film.
aneurysm detected on a routine chest radiograph.

Stephens and Levy [55] reported on a left hilar mass
without pulmonary disease, which was also a pulmonary artery aneurysm. Chung et al [52] have described a case of pulmonary artery aneurysm in
which clinical clues and conventional imaging suggested a lung tumor, and the actual nature of the
lesion was discovered at the time of thoracotomy.
Their case shows the importance of an awareness of
this condition in the formulation of a differential
diagnosis for a lung mass.
On the plain film one cannot always distinguish
between idiopathic dilatation of the pulmonary ar-
teries, pulmonary hypertension, and pulmonary valvular stenosis. MR imaging can evaluate pulmonary
valve hemodynamics [56] and pulmonary anatomy
[45] and help make the diagnosis.
Dilatation of the left atrial appendage segment
The two normal structures that reside within this
area are the left atrial appendage (posteriorly) and the
right ventricular outflow tract (anteriorly). Bulging of
the left atrial appendix is the best single radiographic
sign of left atrial enlargement. This is a more reliable
sign than the more popular double contour within
520
the right border of the heart. Murray et al [57] have

shown that widening of the tracheal bifurcation angle
on chest radiographs is an insensitive and nonspecific
sign of left atrial enlargement. This sign is of little
value in diagnosing left atrial enlargement.
There are some occasions when the left atrial
appendage may bulge without being caused by left
atrial enlargement. This happens in aesthenic females
with narrow anteroposterior diameter of the chest or
with straight back syndrome. The heart is compressed
between the sternum and the spine and causes the left
atrial appendage to pop out. Prominence of the left
atrial appendage can also be seen when the appendage herniates through a partial absence of left pericardium [58] or through a localized cardiac defect
secondary to previous cardiac surgery. Left atrial aneurysms also can be detected on a plain film [59,60],
because chest radiography can give a hint of the
diagnosis with a bulky mass of soft tissue density
appearing adjacent to the left atrial appendage. Mediastinal tumors, such as a thymoma, also can create
this appearance. Cross-sectional images are ideal to
clarify such findings, and allow the detection of other
abnormalities of the left atrium [61 66].
Other causes for enlargement of the left atrial
appendage region are complete absence of the pericardium; enlargement of the right ventricular outflow
tract (as in left-to-right shunt or valvular pulmonic
stenosis); levotranspositon of the great arteries; juxtaposition of the atrial appendages; ventricular tumor;
and several other causes [21].
Abnormalities of left ventricular segment
The left ventricle, except in some complex cardiac
malformations, always forms the lower portion of the
left heart border. It is often difficult to distinguish
between different causes for left ventricular segment
enlargement, because it can be caused by abnormalities in the right side or left side of the heart. Relying
on the plain film can be somewhat inaccurate. Even
when the left ventricle is enlarged, the plain film
findings can be inaccurate predictors, as shown in
numerous publications.
Rose and Stolberg [67] in 1982 discussed the
limited use of the plain chest film in the assessment
of left ventricular structure and function. They summarized their findings as follows. The determination
of radiographic cardiac size as an estimation of the
state of cardiac function is one of the more common
correlations made. Despite the widespread use of
these measurements, the correlation between cardiac
function and radiographic appearance, and the validity of clinical judgments based on this correlation, has
not been fully determined. Similarly, the increment

in left ventricular chamber size necessary to produce
a change in the plain film appearance of the left
ventricle has not been defined. The authors [67]
presented the results of a two observer, blind, retrospective analysis of plain film radiographs of the
chest, and related quantitative left ventricular angiograms, and left ventricular pressure studies performed
on 256 patients. They reported the sensitivity, specificity, predictive value, and accuracy of six previously described plain film measurements of left
ventricular size together with determinations of the
extent of left ventricular volume change necessary to
produce a perceptible change in the plain film radiographic appearance of the left ventricle. The performance of each of these measurements proved to
be disappointing. A sensitivity of 75% was not possible using any of the methods unless the left ventricular chamber volume was more than 66% above
the upper limit of normal.
Other papers also describe the difficulty in interpreting plain film findings related to the left ventricular segment. For example, Ten Cate et al [68] in
1977 reported a case report showing the disparity
between the findings of the plain chest film and the
echocardiogram in a patient with severe coronary
artery disease. Whereas the plain chest film showed
a normal cardiothoracic ratio of 50% indicating normal left ventricular size, the echocardiogram showed
features characteristic of a dilated cardiomyopathic
left ventricle with low amplitude of wall motion and
consequently low ejection fraction. These echocardiographic findings were confirmed at cardiac
catheterization and angiography. Based on this case
report, the authors proposed the use of echo instead
of the plain chest film in the evaluation of left ventricular size in the cardiac patient.
Focal abnormal convexities within the left cardiac
border can be ascribed to left ventricular aneurysm,
left ventricular tumor, pericardial cyst or tumor, left
ventricular diverticulum, and mediastinal or lung
tumor [21].
Common causes of cardiac abnormalities seen on

plain film in older adults
A large variety of cardiac diseases can cause
abnormalities on the plain film, such as valvular heart
disease [69 72], pericardial heart disease [58,73],
myocardial disease [68,74], and congenital heart
disease in adults [70,75,76]. These abnormalities are
followed-up with echocardiography or cross-sectional
images, such as CT or MR imaging. The most

common entities are explored next.
Valvular heart disease
The appearance of valvular heart disease on plain
film has been well studied and described [1,2,72].
These patients are now routinely evaluated with echocardiography, and seldom is the plain film used to
make a diagnosis or to assess the severity or acute
521
or chronic nature of the valvular heart disease. Much

more information can be found in textbooks [2].
Echocardiography is the cross-sectional imaging technique of choice for most patients with valvular heart
disease, with a few exceptions. Valvular function
and anatomy can also be followed-up with MR imaging [77 98].
Left ventricular remodeling after pulmonary autograft aortic valve replacement can be evaluated with
color Doppler echocardiography and MR imaging
Fig. 4. A 53-year-old man with ascending aortic aneurysm and bicuspid aortic valve. (A) Chest radiograph suggests a dilated
ascending thoracic aorta. (B) MR angiogram, frontal view confirms the findings. (C) Black blood MR image, coronal image: the
proximal ascending aorta measured 5.5 cm in cross section. (D) Cine MR image in candy cane view shows the shape of the
ascending aorta and valve leaflets. (E) Cine MR image perpendicular to the aortic valve area shows the bicuspid leaflets.
522
Fig. 4 (continued).
[93]. MR imaging can assess the hemodynamic

effects of pulmonary valve replacement in adults late
after repair of tetralogy of Fallot [92]. Tetralogy of
Fallot with absent pulmonary valve can be evaluated
with MR imaging [96]. MR imaging can quantitate
valvular aortic stenosis [79]. MR imaging can be used
to detect and visualize a bicuspid aortic valve (Fig. 4).
Myocardial disease
Most myocardial diseases are studied with echocardiography, nuclear medicine, or the newer delayed hyperenhancement MR imaging techniques
[99 103]. Plain film findings are not always helpful or suggestive. In the case of ischemic heart there
are many known complications of infarcts, such as
aneurysms and cardiac chamber dilatation [104,105].
Many of these are not discernable on plain films. For
example, left ventricular aneurysms are easily detected and characterized by MR imaging [104,106]
or CT (Fig. 5).
Pericardial disease
Pericardial effusions and congenital absence of
the pericardium have very typical appearances on
the plain film [107,108]. Such findings can be followed-up with any of the cross-sectional techniques
[107 117]. Pericardial masses and cysts are sometimes more difficult to recognize on the plain film,
but have been well described [58,73].
Faridah and Julsrud [109] in 2002 reported that
although much has been published regarding congenital absence of pericardium, it is essential that this
anomaly, like an old friend, be revisited from time to
time. They present a review of this anomaly with
emphasis on its embryologic process. With the advances in MR imaging, absence of pericardium can
now be diagnosed with ease. Gatzoulis et al [110]

in 2000 reported how isolated congenital absence of
the pericardium has a common presentation pattern
with periodic stabbing chest pain mimicking coronary
artery disease. Chest radiograph and MR imaging
are required for definitive diagnosis. Symptomatic
patients with the complete form may benefit from
pericardioplasty. The study by Gatzoulis et al [110]
was based on 10 patients (three males, seven females)
who presented at a median age of 21 years (range,
2 53 years) with paroxysmal stabbing chest pain,
largely nonexertional (9 of 10), and heart murmur
with an abnormal chest radiograph (1 of 10). Three
patients had partial and seven had complete congenital absence of the pericardium (all seven had marked
lateral displacement of the cardiac apex). Chest
radiograph combined with MR imaging were key to
establishing the diagnosis; a tongue of lung tissue
interposing between the main pulmonary artery and
aorta was the most consistent diagnostic feature.
Pericardial effusions also can be followed-up with
MR imaging [118 122]. Chong and Plotnick [121]
in 1995 reviewed how imaging modalities can help
characterize pericardial effusions and tamponade.
Pericardial effusions may be present in a variety of
clinical situations, often presenting challenging clinical diagnostic and therapeutic problems. Although
several imaging modalities are available, echocardiography has become the diagnostic method of choice
because of its portability and wide availability. CT
and MR imaging may also be used and may be more
accurate. A pericardial effusion under pressure may
result in hemodynamic compromise and tamponade.
Although there are several echocardiographic clues to
tamponade (including diastolic chamber collapse,
Doppler flow velocity paradoxus, and inferior vena
cava phlethora), the diagnosis remains clinical and
hemodynamic. Neumann et al [120] in 2002 described how cross-sectional imaging helps to distinguish between paracardial lipodystrophy versus
pericardial effusion in HIV-positive patients. The lipodystrophy syndrome is a side effect of antiretroviral
treatment in HIV-positive patients. Neumann et al
[120] reported a 52-year-old man with HIV, diagnosed 10 years previously, who was being treated
with a combination of nelfinavir, nevirapine, and
stavudine. Echocardiographic examination showed a
low echogenic pericardial space that had increased
from 4 to 18 mm over a 10-month period. The
diagnosis of paracardial adipose tissue was verified
by MR tomography. Misinterpretation of the lipodystrophy as a pericardial effusion and a subsequent
puncture can have serious complications. It is strongly suggested that further differential diagnosis be
523
Fig. 5. (A, B) Frontal and lateral chest radiographs suggest a focal contour anomaly along the left chamber border (frontal) and
bulging of the posterior contour (on lateral). (C, D) Follow-up MR image demonstrated a large left ventricular aneurysm. Both
black blood (C) and bright blood (D) four-chamber views clearly delineate the aneurysm. (E) A coronal black blood MR image
confirms the origin of the left chamber contour anomaly.
524
used for HIV-positive patients with an echocardiographic suspicion of pericardial effusion. Differential
diagnosis by MR tomography is possible.
Adult congenital heart disease
The most important diagnostic features of conventional radiographs in the study of congenital heart
disease have been well described by Grainger [123]
and many others. These diagnostic features are the
pulmonary vasculature; the size of the heart; the
shape of the heart; the position, size, and shape of
the main pulmonary arteries; the position, size, and
shape of the ascending aorta and its arch; the presence
of associated features, (eg, skeletal changes); and
cardiac and visceral situs. Grainger [123] provided
an excellent summary of the importance of the shape
of the heart and how it may be very suggestive
of a specific congenital abnormality. These shapes
are usually described in picturesque and interesting
terms, such as egg-shaped (also called egg lying on
its side, or apple on a string, with a narrow vascular
pedicle) heart of uncorrected transposition of the
great vessels; the sitting-duck heart seen with persistent truncus arteriosus, an elevated rounded cardiac
apex, high right aortic (truncus) arch (in 30% 50%
of cases) and concave pulmonary bay; the bootshaped heart of tetralogy of Fallot, with an elevated
cardiac apex, right aortic arch (in 10% 30% of cases)
and narrow vascular pedicle; the figure-of-eight or
snowman or cottage loaf of bread heart of supracardiac total anomalous pulmonary venous drainage. Epsteins anomaly also presents with a typical
box-shaped square heart, with a prominent right
atrium and the atrialized portion of the right ventricle
(Fig. 6).
As pointed out in 1986 by Grainger [123], now
that corrective surgery is being increasingly practiced,
these picturesque descriptions are less frequently
applicable. Because of corrective surgery the cardiac
chambers and great vessels do not have time to develop the size and shape that produces the characteristic cardiac silhouette of the particular anomaly.
Also, these interesting descriptive shapes only occur
in the minority of examples of each abnormality. A
diagnosis must never be excluded because the shape
of the heart is not characteristic.
Some of the publications addressing this topic are
reviewed next. Baron in 1999 [124] described the
plain film diagnosis of common congenital cardiac
anomalies in the adult. Congenital cardiac lesions in
the adult have characteristic roentgen patterns that
should be recognized by the radiologist. In other
instances, abnormalities in the aorta or the position
of the organs can indicate the likelihood of associated

cardiac anomalies. An increasing number of congenital cardiac patients are surviving into adult life
because of successful treatment. Some of the complications of these repairs can be recognized on
routine chest films. Steiner et al in 1995 [125] reviewed congenital heart disease in the adult patient
and the value of plain film chest radiology. The
authors made the following comments. Congenital
heart disease is a major clinical problem in children,
occurring in 0.8% of newborns. In the past, most
patients with congenital heart disease died in infancy.
With improved surgical and postoperative care, and
more accurate preoperative evaluation, the overall
10-year survival rate is greater than 90%. As a result,
more than 500,000 adults in the United States have
surgically treated congenital heart disease. In addition, at least 150,000 adults are thought to have
unrecognized, misdiagnosed, or recognized but untreated congenital heart disease. Diagnostic imaging
procedures for the evaluation of congenital heart
disease include plain film radiology, fluoroscopy,
angiocardiography, echocardiography, scintigraphy,
CT, and MR imaging. Each has unique and overlapping abilities to characterize precisely cardiovascular anatomy and pathophysiology. Steiner et al
[125] conclude from their 1995 study that the interpreter of the plain film radiograph has a unique
opportunity to identify and often characterize the
severity of a congenital cardiac disorder that may
be unrecognized by the patients physician. Important
clues found on plain film radiographs suggest either
additional studies to pinpoint the type of congenital
heart disease more precisely or that no additional
studies are needed because the recognized lesion is
incidental and not of clinical significance.
MR imaging can be used to evaluate partial
anomalous pulmonary venous return (scimitar syndrome) [126]. Cine MR imaging and three-dimensional contrast-enhanced MR angiography provide
noninvasive diagnostic techniques in the evaluation
of anomalous pulmonary venous return. MR angiography has also been used to evaluate pulmonary
sequestration [127 134]. Zhang et al [129] in 2001
evaluated the feasibility of contrast-enhanced threedimensional MR angiography in identifying the systemic blood supply in pulmonary sequestration in
three patients. Contrast enhanced three-dimensional
MR angiography clearly demonstrated systemic arteries from the descending thoracic aorta supplying the basilar segments of the lower lobe in each
case. Pulmonary veins from these segments, which
drained into the left atrium in two cases, also were
detected. MR angiography images were consistent
525
Fig. 6. A 55-year-old man with Ebsteins anomaly. (A, B) Frontal and lateral chest film show decreased pulmonary vascularity
with cardiomegaly. There is severe right atrial enlargement. (C, D) Coronal and axial bright blood MR images show very
enlarged right atrium, and the associated triscuspid regurgitation (black flow jet). This anomaly involves displacement of the
attachment of the tricuspid leaflet, with atrialization of the right ventricle. (Courtesy of Arthur Stilmann and Richard White,
Cleveland Clinic, Cleveland, OH.)
with those observed in surgery. Ito et al [134] in 2003

reported on a newborn boy with extralobar pulmonary sequestration in the right upper thoracic region.
This was an extremely rare case of extralobar pulmonary sequestration in which anomalous blood
supply from the subclavian artery was seen preoperatively on radiographs.
The plain film also can show an enlarged cardiac
silhouette caused by congenital enlargement of the
right atrium [135]. Congenital malformation of the
right atrium or the coronary sinus is rare, and cases
are classified into one of the following four cate-
gories: (1) congenital enlargement of right atrium,

(2) single diverticulum [136], (3) multiple diverticula
of the right atrium, and (4) diverticulum of the
coronary sinus.
Plain film abnormalities in patients with lung

cancer that suggest cardiac, pericardial, or large
vessel involvement
Patients with lung cancer, either primary or metastatic [137], are almost always followed-up with
526
cross-sectional imaging (MR imaging or CT). For

cancer staging purposes there is the need to evaluate for tumor invasion of the pericardium, vascular
structures, or the heart [138,139]. Although echocardiography can be a first line of defense [140,141],
this is often better evaluated using CT or MR imaging
[142 148].
Tumor invasion can cause pulmonary artery
pseudoaneurysms [149]. Obstruction of the SVC can
cause systemic-to-pulmonary venous shunts [150].
Pulmonary artery obstruction may be caused by
tumor within or external to the arteries [151], and
possibly be noted on plain films. This can cause a
life-threatening compromise in pulmonary flow. Doppler echocardiogram or CT or MR imaging can then
be used to show narrowing of the pulmonary arteries or total occlusion of the pulmonary arteries.
Rarely, small cell lung carcinoma can invade the
left atrium through a pulmonary vein [141]. Crosssectional imaging can help assess the intracavitary
extension of such tumors and the outcome of
therapy (Figs. 7 9). Echocardiography is a relatively low-cost, readily available imaging tool, complementary to radiologic techniques and useful in
evaluating thoracic tumors involving the cardiovascular structures.
Stoblen et al [152] in 1997 evaluated threedimensional CT for the visualization of tumor extent
with respect to the infiltration of the pulmonary
arteries in patients with locally advanced small and
non small cell lung cancer. A total of 61 examinations in 40 patients with bronchial carcinoma were
performed with contrast-enhanced (150 mL injected
with 4 mL/s after bolus tracking) 2-mm spiral
CT (pitch 1.5, increment 1 mm) (Somatom Plus4,
Siemens AG, Germany). Using the implemented

software cine mode, surface shaded display, and
maximum intensity projection, reconstructions of the
pulmonary arteries and the tumors were generated.
Stoblen et al [152] concluded that in comparison
with conventional spiral thoracic CT, the optimized
cross-sectional images of the mediastinum represented a benefit with the technique described, and
this can help to diagnose vessel infiltration and may
be of value for pretherapeutic and posttherapeutic
staging in modern stage-adapted multimodality treatment programs.
Giovagnoni et al [153] in 1992 reported on the
evaluation of the pulmonary artery by cine MR
imaging in 24 patients with mediastinal (12 patients)
and lung (12 patients) neoplasms. These patients
were evaluated with cine MR imaging in addition
to conventional methods (plain radiography, CT,
MR imaging, and endoscopy). Using a 1-T system
and gradient echo (FLASH repetition time 250
350 milliseconds, echo time 12, flip angle 25
60 degrees) multislice pulse sequence, cine MR imaging was performed combined with cardiac gating.
At the level of the pulmonary artery, the CT, spin
echo MR imaging, and cine MR imaging findings
were evaluated blindly by three groups of radiologists
to determine whether the pulmonary artery was
infiltrated or dynamically stenotic. Cine MR imaging
allowed an overall diagnostic accuracy of 100% in
comparison with spin echo MR imaging (91.6%) and
CT (91.6%).
Hasegawa et al [154] in 2003 performed threedimensional dynamic MR imaging on 30 consecutive
patients with primary lung cancer to identify its
usefulness for detecting hilar adenopathy shown at
Fig. 7. A 55-year-old man suspected of having a massive pulmonary embolism. (A) Initial chest radiograph was unremarkable.
(B) Follow-up CT scan revealed a large intracardiac tumor extending in the right atrium and part of right ventricle. (Courtesy of
Francisco Garcia-Morales, VA North Texas Healthcare Center, Dallas, TX.)
527
Fig. 8. A 59-year-old man with metastatic lung cancer (non small cell carcinoma, stage T4) presents with bilateral lower
extremity edema and right upper extremity edema and skin rash. (A) Chest radiograph shows the right upper lung lobe mass.
(B) CT scan shows cross section of mass. The SVC is displaced anteriorly. (C) Three-dimensional contrast-enhanced MR
angiogram shows the arterial and venous structures, and demonstrates mild compression but patency of the SVC. (D) A
noncontrast time-of-flight MR venogram shows only the venous structures, and confirms the same findings.
surgery. They conclude that hilar adenopathy on

three-dimensional dynamic MR imaging correlated
well with that of surgical findings on patients with
primary lung cancer. It may have the potential to
make an accurate preoperative evaluation of hilar
lymph node metastasis from lung cancer.
Ohno et al [155] in 2001 reported on the use of
multiphase ECG-triggered three-dimensional contrast-enhanced MR angiography for evaluation of
hilar and mediastinal invasion of bronchogenic carcinoma. The purpose of their study was to evaluate
the usefulness of cardiac synchronized MR angiography [155] for improving image quality and
detection of hilar and mediastinal invasion of bronchogenic carcinoma. Fifty patients suspected of
having hilar or mediastinal invasion of bronchogenic
carcinoma underwent contrast-enhanced CT and MR

imaging including conventional and ECG-triggered
MR angiography. Twenty patients subsequently also
underwent surgical resection. Vascular enhancementto-background ratio, vascular enhancement-to-tumor
ratio, signal-to-noise ratio, contrast-to-noise ratio, and
image quality scores of thoracic vessels obtained with
both MR angiography techniques were determined
and compared. In addition, the diagnostic accuracy
of tumor invasion of pulmonary vessels was compared. Vascular enhancement-to-background ratios
and vascular enhancement-to-tumor ratios of both
MR angiography techniques were not significantly
different. ECG-triggered MR angiography significantly improved signal-to-noise ratios and contrastto-noise ratios (P < .05). Two readers judged that
528
Fig. 9. A 51-year-old with metastatic renal cell carcinoma (pulmonary mets). (A) Chest radiograph shows multiple nodular
masses. (B, C) Transaxial MR images confirm the multiple lung masses, and suggest that some may be invading the pericardium.
(D) Close up of an MR image showing one mass adjacent to the left posterior pericardium. (E, F) Cine MR images shown at
two points in the cardiac cycle demonstrate how the mass seen in Fig. 9D does not invade the pericardium. Myocardial
tagging during a MR image study (not shown here) can further improve the visualization of possible pericardial involvement
and adhesions.
overall image quality of ECG-triggered MR angiography was better than that of conventional MR
angiography (kappa 0.41). The authors concluded
that ECG-triggered MR angiography improves the
image quality and the detection of hilar and mediastinal invasion of bronchogenic carcinoma.
Edmondstone [156] in 1998 reported on flitting
radiographic shadows as an unusual presentation of
cancer in the lungs. The author states that tumor
involvement of pulmonary blood vessels occurs frequently in advanced lung cancer and occasionally
may cause pulmonary infarction. This author also
reports a case of diffuse obstruction of pulmonary
arteries by cancer in which no primary tumor was
found, and which presented as flitting radiographic
opacities because of pulmonary infarction.
Takahashi et al [157] in 2000 also evaluated the
ability of breath-hold gadolinium-enhanced threedimensional MR angiography to assess the invasion
of the pulmonary vein and the left atrium by lung cancer in 20 consecutive patients with lung cancer. They
concluded that breath-hold gadolinium-enhanced
three-dimensional MR angiography is suitable for assessing invasion of the pulmonary vein and the left
atrium by lung cancer.
Neither MR imaging nor CT, however, is always
perfect in predicting tumor resectability. Because of
discrepancies in predicting resectability by imaging
techniques (CT and MR imaging) compared with
actual intraoperative findings Loscertales et al [158]
in 2002 reported on how they perform systematic
exploratory videothoracoscopy as the first step in the
surgical evaluation of patients with lung cancer. The
authors [158] claim that resectability of centrally
located primary tumors with intrapericardial extension (clinical T4) can only be established by direct
examination of the pericardial sac contents. In these
instances, they added videopericardioscopy to their
presurgical evaluation protocol. Their study suggests
that exploratory videothoracoscopy is superior to imaging techniques (CT or MR imaging) in detecting
tumor extension into the pericardium. In addition,
short of an exploratory thoracotomy, videopericardioscopy seems to be the most definitive study to establish resectability of centrally located tumors with
pericardial invasion. Unnecessary exploratory thoracotomies can be avoided.
Postsurgical and posttraumatic abnormalities seen

on plain films
Patients who undergo surgical repair of congenital
heart disease, thoracic aortic disease, or who undergo
529
coronary artery bypass grafting or stent placement for

ischemic heart disease, can have postsurgical complications with unusual appearances on the plain film.
Most common is the case of a patient with repair of
an aortic dissection or aneurysm (Fig. 10). After
major cardiovascular surgery patients typically undergo periodic follow-up with cross-sectional imaging [159,160]. Recognition of sudden changes in the
appearance of the plain film after thoracic aortic
surgery is very important and should prompt an
immediate follow-up MR imaging or CT study.
Saphenous venous bypass graft aneurysm can present
as mediastinal or pericardiac masses on the plain film
[161 166].
Fattori et al [167] in 1999 assessed the value of
MR imaging in the detection of postoperative complications after composite valve graft replacement in
52 patients. Normal postoperative perigraft thickening (V 10 mm) was observed in 42 patients. Ten
patients had abnormal periprosthetic thickening of
15 to 52 mm. Gadolinium-enhanced MR imaging
demonstrated leakage in 5 of those 10 patients. The
lack of enhancement excluded the presence of bleeding in the remaining five patients (three with chronic
hematomas, one with infection, and one with granulation tissue). These findings were confirmed at
surgery or with subsequent follow-up MR imaging
examinations. Fattori et al [167] concluded that
MR imaging was an optimal imaging modality for
evaluating the morphologic characteristics of composite grafts and reimplanted coronary arteries. Gadolinium-enhanced MR imaging is a simple, accurate,
and noninvasive method for detecting a leak, which
necessitates urgent repeat surgery. An example of
such perigraft leak is shown in Fig. 11.
Nguyen and Nguyen [168] have reported on the
plain film findings with foreign bodies in the pericardial sac. They report two cases of gunshot wounds
of the chest, in each of which a bullet was retained in
the pericardial sac. Because the bullets appear to be
out of focus, one may think they are in the cardiac
chambers, but the images of bullets retained in the
pericardial sac may also be blurred because of the
spinning effect of the heartbeat on the bullet.
Fultz et al [35] in 1998 evaluated chest radiographic features of nontraumatic mediastinal hemorrhage occurring after extrapericardial thoracic aorta
rupture. Twenty-seven consecutive chest radiographs
obtained at admission of patients with hemorrhage
from ruptured thoracic aorta aneurysms, aortic dissections, or penetrating aortic ulcers were randomized
with radiographs of 23 subjects with nonruptured
thoracic aorta aneurysms, 20 subjects with nonruptured dissections, and 20 control subjects. Logistic
530
regression analysis showed a combination of obscuration or convexity of the aorticopulmonary window and a displaced left paraspinal interface to be
the most useful predictor of hemorrhage (P < .05).
Rank correlation analysis indicated obscuration or
convexity of the aorticopulmonary window; a displaced left paraspinal interface; enlarged aortic knob
width; enlarged thoracic aorta size; an enlarged,
obscured, or irregular aortic margin; and left pleural
or extrapleural space fluid were potential individual
predictors of hemorrhage (P < .05). Observer sensitivities for recognizing hemorrhage were 30% to
59% and specificities were 83% to 91%. Sensitivities
for distinguishing an abnormal (N = 70) from a
normal (N = 20) mediastinum were 79% to 90%
and specificities were 65% to 90%. Fultz et al [35]
concluded that obscuration or convexity of the aorticopulmonary window and a displaced left paraspinal
interface on radiographs may indicate mediastinal
hemorrhage. Further imaging is required to establish
a definitive diagnosis.
Plewa et al [169] in 1997 studied cervical prevertebral soft tissue measurements and chest radiographic findings in acute traumatic aortic injury.
Mediastinal widening, aortopulmonic window opacification, and blurring of the aortic knob were the most
sensitive chest radiography findings in acute traumatic aortic injury, although each of these lacked
useful specificity and accuracy. Cervical soft tissue
swelling is not a useful marker for acute traumatic
aortic injury.
Kram et al [32] in 1989 performed a 10-year
retrospective analysis of 82 patients with suspected
thoracic aortic rupture caused by blunt chest trauma
to define which symptoms and signs were helpful in
making an early diagnosis. Chest roentgenographic
signs seen with significantly greater frequency in the
12 patients with thoracic aortic rupture than in
70 patients without such rupture included a widened
paratracheal stripe (seven patients); deviation of the
nasogastric tube or central venous pressure line
(five patients); blurring of the aortic knob (nine
patients); abnormal paraspinous stripe (six patients);
and rightward tracheal deviation (five patients).
Mediastinal widening of greater than 8 cm occurred
in 11 of the 12 patients with thoracic aortic rupture
(sensitivity, 92%); its specificity, however, was only
10% (11 true-positive and 63 false-positive results).
531
In patients in hemodynamically stable condition who

display these findings, immediate aortography should
be considered.
Savastano et al [34] in 1989 looked at the value of
plain chest film in predicting traumatic aortic rupture.
Plain chest film performed after blunt chest trauma
showed blurring of the left pulmonary hilum in
53% of cases of traumatic aortic rupture (group A,
N = 15), and in no cases with negative aortography
(group B, N = 10). According to these authors, this
sign can be explained by diffusion of mediastinal
hemorrhage through the peribronchovascular connective tissue; the close relationship between aortic
isthmus, the side of most frequent rupture, and the
left pulmonary hilum is the anatomic basis for this
asymmetric finding. Review of all radiologic alterations of the plain chest film showed a statistically
significant difference between groups A and B only
for mediastinal widening and aortic knob alterations
(P < .05). A combination of some findings caused
by aortic injury (mediastinal widening, aortic knob alterations, shift of trachea and left main bronchus, left
apical cap, left hilar blurring, obscuring of descending aorta) was typical of aortic rupture when four or
more signs were found (33%); these findings were
absent only in patients with negative aortogram (8%).
In the remaining cases (60%), the plain chest film
showed two or three of these signs in both groups,
making it impossible to differentiate between patients
with and without aortic injury.
Jagannath et al [31] in 1986 described how findings on plain chest radiographs of patients with aortic
dissection are variable and often overlap those of
patients without dissection. To determine which findings were most useful in predicting aortic dissection,
plain chest radiographs from 36 patients with aortographically proved aortic dissection and 36 patients
from a control population were randomized and
analyzed independently by five radiologists for the
presence of various radiographic features associated
with this condition. A widened aortic knob, widened
descending aorta, and widened mediastinum showed
the greatest interobserver agreement (P < .001),
although the overall interobserver agreement was
poor. The final conclusion of the radiologists was a
better predictor of dissection than any of the individual radiographic features alone. Widening of the
mediastinum (P < .001) and widening of the aortic
Fig. 10. A 55-year-old man with prior thoracic aortic aneurysm repair. (A, B) Frontal and lateral chest radiographs suggest a
tortuous aorta, with prominence of the aortic knob. (C, D) Oblique sagittal (candy cane) views of the aorta during a threedimensional contrast enhanced MR angiogram demonstrate the complex postsurgical appearance of the thoracic aorta, with
partial dissection, aneurismal dilatation, and thrombus formation in the residual false lumen. (E) Transaxial cine MR image
demonstrates dissection in the ascending aorta.
532
Fig. 11. A 33-year-old man with Marfan syndrome identified at age 27 when he was found to have an aortic dissection. He
underwent surgery at age 29 involving aortic valve replacement and repair of ascending aortic aneurysm. Patient presented with
new chest pain. (A) Chest radiograph does not show any significant aortic anomaly; the cardiac silhouette is at upper limits of
normal, with a prominent arch but no widening of the mediastinum. (B, C) Two views from an MR angiogram (B, frontal, and C,
candy-cane) show the graft repair of the ascending aorta, but no aneurysm or dissection. (D) CT scan shows aneurysmal
dilatation of the ascending aorta (diameter 7 cm), starting at the aortic root and extending to the proximal arch (not shown), with
contrast accumulation in the perigraft area, suggesting a postsurgical leak. (E, F) MR image cross-sectional imaging also shows
uptake of contrast in the peri-graft area. The appearance of the perigraft area after surgical repair of ascending thoracic aorta can
be quite variable.
533
Fig. 11 (continued).
knob (P < .012) were the only two radiographic

features of significance in predicting dissection. In a
stepwise multiple logistic regression model, the radiologists achieved an overall accuracy of 85%, a
sensitivity of 81%, and a specificity of 89%. Although this study illustrates the usefulness of plain
chest radiographs in diagnosing aortic dissection,
there was poor interobserver agreement.
With the advent of endoluminal repair of aortic
disease the importance of periodic cross-sectional
imaging with MR imaging or CT and MR angiography or CT angiography has even increased further
[4,93,160,170 175].
Other uses of MR imaging and CT in the thorax

Cross-sectional imaging and correlation of the
plain films is often needed for anomalies not directly
related to the heart or great vessels. These include
pulmonary veins, systemic veins, lung masses, mediastinal masses, the pleura, chest wall, and the
thoracoabdominal junction. These areas are only
briefly reviewed here, using mostly articles published
as part of the MR Imaging of the Thorax, February
2000 issue of the Magnetic Resonance Imaging
Clinics of North America.
White [176] described how MR imaging can be
used to image thoracic vein abnormalities. These
thoracic venous anomalies can be classified conveniently as systemic or pulmonary. Congenital systemic venous anomalies, such as anomalies of SVC
or anomalies of the inferior vena cava, are often incidental findings. Congenital pulmonary venous
anomalies, however, such as total or partial anomalous pulmonary venous return [126], and other
anomalies, such as bronchopulmonary sequestration [128], are more likely to manifest with cyanosis
and to be associated with congenital cardiac anomalies, especially atrial septal defects. Thrombosis,
tumor invasion, and inflammatory conditions often
also cause acquired systemic and pulmonary venous
anomalies (Fig. 12). MR imaging provides excellent
delineation of the abnormal vessels and associated
lesions. Both cross-sectional imaging and functional
imaging using flow measurement or MR angiography can be very useful in delineating these abnormalities. Plain film findings sometimes give a clue
as to a particular type of abnormality, such as in
the case of certain types of anomalous pulmonary
venous connections.
Pulmonary vein ablation offers the potential to
cure patients with atrial fibrillation. Cross-sectional
imaging is routinely used to investigate the incidence
of pulmonary vein stenosis after radiofrequency
catheter ablation of refractory atrial fibrillation
[177 182]. Arentz et al [178] reported on this in
2003 and concluded that at 2-year follow-up, the risk
of significant pulmonary vein stenosis or occlusion
after radiofrequency catheter ablation of refractory
atrial fibrillation with conventional mapping and
ablation technology was 28%. Distal ablations inside
smaller pulmonary veins should be avoided because
of the higher risk of stenosis than ablation at the
ostium. Dill et al [180] in 2003 investigated the
incidence and time course of pulmonary vein stenosis
after radiofrequency catheter ablation within a period
of 3 months. Contrast-enhanced MR angiography
was used to visualize pulmonary veins and was
compared with radiographic angiography. Dill et al
[180] conclude that the occurrence and progression
of pulmonary vein stenosis is a potential significant
534
Fig. 12. A 56-year-old with left upper extremity swelling. (A) Chest radiograph shows cardiomegaly with pulmonary congestion
and bilateral effusions. No left apical mass is noted. (B, C) Frontal views of contrast-enhanced MR venograms show the upper
thoracic and neck veins. The angiogram was performed twice, first with contrast injection in the right arm (B) and then in the left
arm (C). The MR venogram obtained with the left-sided contrast injection demonstrates severe narrowing of a long segment of
the left inominate vein (C) because of prior instrumentation.
complication of radiofrequency catheter ablation in

the orifice of pulmonary veins. These findings may
have an impact on the technical performance of this
intervention. In addition, long-term studies are necessary to evaluate lumen reduction over time. MR
angiography is a noninvasive, reproducible imaging
modality for this purpose. Kato et al [179] in 2003
reported on the benefits of preprocedural MR imaging of pulmonary veins including the ability to
evaluate the number, size, and shape of the pulmonary veins. MR imaging also provides an assessment
of the severity of pulmonary vein stenosis.
Superior vena cava obstruction causes altered
flow dynamics by collateral pathways, and can cause
abnormal hepatic enhancement patterns [183 186].

CT and MR imaging are ideal to evaluate these
collateral pathways.
The correlation of MR imaging and the plain
film for the delineation of lung cancer and thoracic
lymph nodes is probably not as well established
[187,188]. CT cross-sectional imaging remains the
modality of choice in combination with positron
emission tomography imaging. Technology is improving, however, and MR imaging might end up
helping to characterize lymph nodes. MR imaging of
mediastinal lymph nodes can be improved using a
superparamagnetic contrast agent [189]. Respiratorytriggered short inversion time inversion recovery
turbo spin echo MR imaging has been used with

success to differentiate between metastatic and nonmetastatic lymph nodes in patients with non small
cell lung cancer [190].
Mediastinal masses and abnormalities are very
well delineated by either CT or MR imaging [191].
MR imaging provides the additional ability to acquire
images directly in multiple planes. Both MR imaging
and CT are used to provide additional information
about the location and extent of the disease. MR
imaging can be used to confirm the cystic nature of
mediastinal lesions that appear solid on CT. MR
imaging is the modality of choice for imaging neurogenic tumors because it can optimally demonstrate
the number and nature of lesions, intraspinal extension, and craniocaudad extent. MR imaging also is
especially useful for evaluating the mediastinum of
patients in whom the administration of iodinated
contrast material is contraindicated.
MR imaging also is an excellent technique to
evaluate the pleura, chest wall [192], and the thoracoabdominal junction [193]. The ability to acquire
multiplanar imaging planes is very useful when
imaging abnormalities of the diaphragm. Abnormalities involving the diaphragm can be difficult to
characterize because of its complex shape and contour, often providing a challenge for the radiologist.
Many processes related to adjacent organs, such as
liver, pleura, and lungs, could involve this region
secondarily. MR imaging is also an excellent technique to correlate with plain film findings to characterize better benign and malignant chest wall lesions.
Specifically, lesions in the apex of the lungs are better
visualized and characterized with multiplanar crosssectional imaging.
Summary
Multiplanar imaging using MR imaging or CT
offers significant added information when trying to
clarify abnormalities seen on a plain film of the chest.
Knowledge of the plain film appearance of the
normal heart is an essential starting point. The choice
between MR imaging or CT or MR angiography or
CT angiography as the most appropriate follow-up
study is in a state of flux, because technology is
changing rapidly and the use of new scanners and
postprocessing techniques is proliferating. The actual
selection of one cross-sectional modality over the
other seems to be dictated more by availability of
scanners and personal choice, besides generic concerns about radiation dose [194] and the use of
iodinated contrast media. It is hoped that this article
535
helps the reader better appreciate the correlation

between cross-sectional MR imaging and plain
film abnormalities.
Acknowledgments
The author thanks Murray G. Baron, MD, from
Emory University Hospital in Atlanta, GA, for
providing his opinions and insight in the evaluation
of plain film findings. Portions of this manuscript are
based on his 2001 review in the International Journal
of Cardiovascular Imaging [3]. The author also
thanks his colleagues for suggesting or providing
some of the cases illustrated here: Francisco GarciaMorales, MD, from the VA North Texas Healthcare
System, Dallas, Texas; Arthur Stillman, MD, PhD,
and Richard White, MD, from the Cleveland Clinic,
Cleveland, Ohio.
References
[1] Higgins CB. Essentials of cardiac radiology and imaging. Philadelphia: JB Lippincott; 1992.
[2] Miller SW. Cardiac radiology: the requisites. St Louis:
Mosby; 1996.
[3] Baron MG. Correlation of plain films and MR of the
heart. Int J Cardiovasc Imaging 2001;17:453 6.
[4] Weigel S, Tombach B, et al. Thoracic aortic stent
graft: comparison of contrast-enhanced MR angiography and CT angiography in the follow-up: initial
results. Eur Radiol 2003;13:1628 34.
[5] Pereles FS, McCarthy RM, et al. Thoracic aortic dissection and aneurysm: evaluation with nonenhanced
true FISP MR angiography in less than 4 minutes.
Radiology 2002;223:270 4.
[6] Kreitner KF, Kunz RP, et al. Contrast-enhanced threedimensional MR angiography of the thoracic aorta:
experiences after 118 examinations with a standard
dose contrast administration and different injection
protocols. Eur Radiol 2001;11:1355 63.
[7] Silverman JM, Raissi S, et al. Phase-contrast cine MR
angiography detection of thoracic aortic dissection.
Int J Cardiovasc Imaging 2000;16:461 70.
[8] Krinsky GA, Reuss PM, et al. Thoracic aorta: comparison of single-dose breath-hold and double-dose
non-breath-hold gadolinium-enhanced three-dimensional MR angiography. AJR Am J Roentgenol
1999;73:145 50.
[9] Rubin GD. CT angiography of the thoracic aorta.
[10] Lawler LP, Fishman EK. Multi-detector row CT of
thoracic disease with emphasis on 3D volume rendering and CT angiography. Radiographics 2001;21:
1257 73.
[11] Batra P, Bigoni B, et al. Pitfalls in the diagnosis of
536
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]

thoracic aortic dissection at CT angiography. Radiographics 2000;20:309 20.
van den Hout RJ, Lamb HJ, et al. Real-time MR
imaging of aortic flow: influence of breathing on left
ventricular stroke volume in chronic obstructive pulmonary disease. Radiology 2003;229:513 9.
Hori Y, Yamada N, et al. Rapid evaluation of right
and left ventricular function and mass using real-time
true-FISP cine MR imaging without breath-hold:
comparison with segmented true-FISP cine MR imaging with breath-hold. J Cardiovasc Magn Reson
2003;5:439 50.
Wintersperger BJ, Nikolaou K, et al. Single breathhold real-time cine MR imaging: improved temporal
resolution using generalized autocalibrating partially
parallel acquisition (GRAPPA) algorithm. Eur Radiol
2003;13:1931 6.
Nayak KS, Hu BS. Triggered real-time MRI and
cardiac applications. Magn Reson Med 2003;49:
188 92.
Saito K, Saito M, et al. Real-time four-dimensional
imaging of the heart with multi-detector row CT.
Radiographics 2003;23:e8.
Kazerooni EA, Collins J, et al. A curriculum in chest
radiology for diagnostic radiology residency, with
goals and objectives. Acad Radiol 2000;7:730 43.
Collins J, Reddy GP, et al. A curriculum in cardiothoracic radiology for medical students, with goals
and objectives. Acad Radiol 2001;8:1247 51.
Boxt LM, Reagan K, et al. Normal plain film examination of the heart and great arteries in the adult.
J Thorac Imaging 1994;9:208 18.
Boxt LM. Plain-film examination of the normal heart.
Higgins CB. Radiography of acquired heart disease.
In: Higgins CB, editor. Essentials of cardiac radiology and imaging. Philadelphia: JB Lippincott Company; 1992. p. 1 48.
Chaumoitre K, Zappa M, et al. Rupture of the right
atrium-superior vena cava junction from blunt thoracic trauma: helical CT diagnosis. AJR Am J Roentgenol 1997;169:1753.
Thorogood SV, Maskell GF. Case report: intravascular lipoma of the superior vena cava CT and MRI
appearances. Br J Radiol 1996;69:963 4.
Shivkumar K, Ravi K, et al. Chest radiographs fail to
detect right ventricular enlargement and right atrial
enlargement in patients with a pure restrictive ventilatory impairment. Chest 1994;106:381 4.
Baron MG. Radiologic notes in cardiology: obscuration of the aortic knob in coarctation of the aorta.
Circulation 1971;43:311 6.
Bell MF, Jernigan TP, et al. Prognostic significance of
calcification of the aortic knob visualized radiographically. Am J Cardiol 1964;13:640 4.
Bell MF, Schaaf RS, et al. The prognostic import of
calcification of the aortic knob. Trans Assoc Life
Insur Med Dir Am 1964;47:33 43.
Chapman I, Kwun C, et al. Utilization of aortic knob
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
calcification for radiographic localization. Exp Med

Surg 1961;19:223 9.
Chapman I. Anatomic and clinical significance of calcification of the aortic knob visualized radiographically. Am J Cardiol 1960;6:281 6.
Symeonidis G, Papanas N, et al. Gravity of aortic
arch calcification as evaluated in adult Greek patients.
Int Angiol 2002;21:233 6.
Jagannath AS, Sos TA, et al. Aortic dissection: a
statistical analysis of the usefulness of plain chest
radiographic findings. AJR Am J Roentgenol 1986;
147:1123 6.
Kram HB, Appel PL, et al. Diagnosis of traumatic
thoracic aortic rupture: a 10-year retrospective analysis. Ann Thorac Surg 1989;47:282 6.
Kram HB, Wohlmuth DA, et al. Clinical and radiographic indications for aortography in blunt chest
trauma. J Vasc Surg 1987;6:168 76.
Savastano S, Feltrin GP, et al. Value of plain chest
film in predicting traumatic aortic rupture. Ann
Radiol (Paris) 1989;32:196 200.
Fultz PJ, Melville D, et al. Nontraumatic rupture of
the thoracic aorta: chest radiographic features of an
often unrecognized condition. AJR Am J Roentgenol
1998;171:351 7.
van Beek EJ, Wild JM, et al. MRI for the diagnosis of
pulmonary embolism. J Magn Reson Imaging 2003;
18:627 40.
Patel S, Kazerooni EA, et al. Pulmonary embolism:
optimization of small pulmonary artery visualization
at multi-detector row CT. Radiology 2003;227:
455 60.
Ley S, Kreitner KF, et al. Bronchopulmonary
shunts in patients with chronic thromboembolic
pulmonary hypertension: evaluation with helical CT
and MR imaging. AJR Am J Roentgenol 2002;
179:1209 15.
Arents Jr DN. Chronic thromboembolic pulmonary
hypertension. N Engl J Med 2002;346:866.
Kruger S, Haage P, et al. Diagnosis of pulmonary
arterial hypertension and pulmonary embolism with
magnetic resonance angiography. Chest 2001;120:
1556 61.
Koito H, Yutaka H. CT and MRI findings of pulmonary hypertension. Nippon Rinsho 2001;59:
1107 12.
Laffon E, Laurent F, et al. Noninvasive assessment of
pulmonary arterial hypertension by MR phase-mapping method. J Appl Physiol 2001;90:2197 202.
Fayad ZA, Ferrari VA, et al. Right ventricular regional function using MR tagging: normals versus
chronic pulmonary hypertension. Magn Reson Med
1998;39:116 23.
Bergin CJ, Hauschildt J, et al. Accuracy of MR angiography compared with radionuclide scanning in
identifying the cause of pulmonary arterial hypertension. AJR Am J Roentgenol 1997;168:1549 55.
Nollen GJ, van Schijndel KE, et al. Magnetic resonance imaging of the main pulmonary artery: reliable
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
assessment of dimensions in Marfan patients on a

simple axial spin echo image. Int J Cardiovasc Imaging 2003;19:141 7 [discussion: 149 50].
Vazquez Munoz E, Rios Blanco JJ, et al. Diagnosis
with helical CT and echocardiography of patients
with suspected chronic thromboembolic pulmonary
hypertension. Ann Med Interne (Paris) 2003;20:
363 6.
Pitton MB, Kemmerich G, et al. Chronic thromboembolic pulmonary hypertension: diagnostic impact
of Multislice-CT and selective Pulmonary-DSA. Rofo
Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 2002;
174:474 9.
Murray TI, Boxt LM, et al. Estimation of pulmonary
artery pressure in patients with primary pulmonary
hypertension by quantitative analysis of magnetic resonance images. J Thorac Imaging 1994;9:198 204.
Bossone E, Duong-Wagner TH, et al. Echocardiographic features of primary pulmonary hypertension.
J Am Soc Echocardiogr 1999;12:655 62.
Laffon E, Vallet C, et al. A computed method for
noninvasive MRI assessment of pulmonary arterial
hypertension. J Appl Physiol 2004;96:463 8.
Agarwal S, Chowdhury UK, et al. Isolated idiopathic
pulmonary artery aneurysm. Asian Cardiovasc Thorac
Ann 2002;10:167 9.
Chung CW, Doherty JU, et al. Pulmonary artery
aneurysm presenting as a lung mass. Chest 1995;
108:1164 6.
Janssens F, Verswijvel G, et al. Proximal pulmonary
artery aneurysm. JBR-BTR 2003;86:83 5.
Oliver TB, Stevenson AJ, et al. Pulmonary artery
pseudoaneurysm due to bronchial carcinoma. Br J
Radiol 1997;70:950 1.
Stephens MG, Levy JM. Left hilar mass without pulmonary disease: pulmonary artery aneurysm. West J
Med 1990;152:93 4.
Reid SA, Walker PG, et al. The quantification of
pulmonary valve haemodynamics using MRI. Int J
Cardiovasc Imaging 2002;18:217 25.
Murray JG, Brown AL, et al. Widening of the tracheal
bifurcation on chest radiographs: value as a sign of
left atrial enlargement. AJR Am J Roentgenol 1995;
164:1089 92.
Rozenshtein A, Boxt LM. Plain-film diagnosis
of pericardial disease. Semin Roentgenol 1999;34:
195 204.
Hoffmann U, Hamed N, et al. Radiological signs of
a left atrial aneurysm. Eur Radiol 2000;10:1332 4.
Morales JM, Patel SG, et al. Left atrial aneurysm.
Ann Thorac Surg 2001;71:719 22.
Ohyama H, Hosomi N, et al. Comparison of magnetic
resonance imaging and transesophageal echocardiography in detection of thrombus in the left atrial
appendage. Stroke 2003;34:2436 9.
Oshima H, Hara M, et al. Cardiac hemangioma of the
left atrial appendage: CT and MR findings. J Thorac
Imaging 2003;18:204 6.
Alam G, Addo F, et al. Detection of left atrial append-
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]
[79]
[80]
[81]
537
age thrombus by spiral CT scan. Echocardiography

2003;20:99 100.
Stollberger C, Ernst G, et al. Left atrial appendage
morphology: comparison of transesophageal images
and postmortem casts. Z Kardiol 2003;92:303 8.
Krishnamoorthy KM. Aneurysm of the left atrial
appendage. Cardiology 2002;97:233 4.
von der Recke G, Schmidt H, et al. Transesophageal
contrast echocardiography distinguishes a left atrial
appendage thrombus from spontaneous echo contrast.
Echocardiography 2002;19:343 4.
Rose CP, Stolberg HO. The limited utility of the plain
chest film in the assessment of left ventricular structure and function. Invest Radiol 1982;17:139 44.
Ten Cate FJ, Vletter WB, et al. The plain chest film as
an unreliable method to determine left ventricular
size. Jpn Heart J 1977;18:878 82.
Stein PD. Assessment of calcific aortic stenosis by
measurement of area circumscribed by calcium on
plain film orifice-view roentgenograms. Chest 1974;
65:518 21.
Spindola-Franco H, Fish BG, et al. Recognition of
bicuspid aortic valve by plain film calcification.
AJR Am J Roentgenol 1982;139:867 72.
Rubin SA, Hightower CW, et al. Giant right atrium
after mitral valve replacement: plain film findings
in 15 patients. AJR Am J Roentgenol 1987;149:
257 60.
Kurtzman RS. Acquired lesions of the mitral valve:
plain film analysis. Semin Roentgenol 1979;14:
131 43.
Lane Jr EJ, Carsky EW. Epicardial fat: Lateral plain
film analysis in normals and in pericardial effusion.
Radiology 1968;91:1 5.
Hammermeister KE, Chikos PM, et al. Relationship
of cardiothoracic ratio and plain film heart volume to
late survival. Circulation 1979;59:89 95.
Bream PR, Elliott LP, et al. Plain film findings of
anatomically corrected malposition: its association
with juxtaposition of the atrial appendages and right
aortic arch. Radiology 1978;126:589 95.
Kurlander GJ, Petry EL, et al. Plain film diagnosis
of congenital heart disease in the newborn period.
Am J Roentgenol Radium Ther Nucl Med 1968;
103:66 77.
Vrachliotis TG, Bis KG, et al. Aorto-mitral valvular
evaluation with MRI using the left parasternal long
axis (angled-vertical long axis) plane. Int J Cardiovasc Imaging 2002;18:61 6.
Colletti PM, DeFrance A, et al. Cardiac MRI cine and
color Doppler in valvular disease: correlative imaging. Magn Reson Imaging 1991;9:343 7.
Friedrich MG, Schulz-Menger J, et al. Quantification
of valvular aortic stenosis by magnetic resonance imaging. Am Heart J 2002;144:329 34.
Rozenshtein A, Boxt LM. Computed tomography and
magnetic resonance imaging of patients with valvular
heart disease. J Thorac Imaging 2000;15:252 64.
Sondergaard L, Stahlberg F, et al. Magnetic resonance
538
[82]
[83]
[84]
[85]
[86]
[87]
[88]
[89]
[90]
[91]
[92]
[93]
[94]
[95]
[96]

imaging of valvular heart disease. J Magn Reson Imaging 1999;10:627 38.
Bommer W. Magnetic resonance and other imaging
techniques in valvular and cardiac disease. Curr Opin
Cardiol 1992;7:216 21.
Schmidt M, Crnac J, et al. Magnetic resonance imaging in valvular heart disease. Int J Cardiovasc Imaging 1997;13:219 31.
Globits S, Higgins CB. Assessment of valvular heart
disease by magnetic resonance imaging. Am Heart J
1995;129:369 81.
Une F. Nuclear magnetic resonance image and CT
scan studies on silent strokes in non-valvular atrial
fibrillation (NVAF). Nippon Rinsho 1993;51:874 8.
Ohnishi S, Fukui S, et al. Assessment of valvular
regurgitation using cine magnetic resonance imaging
coupled with phase compensation technique: comparison with Doppler color flow mapping. Angiology
1992;43:913 24.
Higgins CB, Wagner S, et al. Evaluation of valvular heart disease with cine gradient echo magnetic
resonance imaging. Circulation 1991;84(Suppl 3):
I198 207.
Wagner S, Auffermann W, et al. Diagnostic accuracy
and estimation of the severity of valvular regurgitation from the signal void on cine magnetic resonance
images. Am Heart J 1989;118:760 7.
Germain P, Baruthio J, et al. Value of cine magnetic
resonance imaging in the diagnosis and quantification
of valvular regurgitation: comparison with angiography and Doppler echocardiography. Arch Mal Coeur
Vaiss 1989;82:1659 68.
Onishi S, Fukui S, et al. Clinical evaluation of regurgitant blood flow by rapid cine magnetic resonance
imaging in patients with valvular heart disease. J Cardiol 1989;19:571 82.
Makeeva TV. Extraction of non-magnetic foreign
bodies from deep layers of the cornea by means of
valvular non-continuous trepanation. Zdravookhranenie Kazakhstana 1962;22:22 4.
Vliegen HW, van Straten A, et al. Magnetic resonance imaging to assess the hemodynamic effects of
pulmonary valve replacement in adults late after repair of tetralogy of Fallot. Circulation 2002;106:
1703 7.
Schmid FX, Keyser A, et al. Left ventricular remodeling after pulmonary autograft aortic valve
replacement: evaluation with color-Doppler echocardiography and magnetic resonance imaging. Artif
Organs 2002;26:444 8.
Kuehne T, Saeed M, et al. Sequential magnetic resonance monitoring of pulmonary flow with endovascular stents placed across the pulmonary valve in
growing swine. Circulation 2001;104:2363 8.
Cabrera Duro A, Gilbert Valencia J, et al. Magnetic
resonance in a case of a newborn with pulmonary
valve agenesis and deletion of chromosomal region
22q 11,2. An Esp Pediatr 2000;52:485 7.
Kakizawa H, Ohno T, et al. Tetralogy of Fallot with
[97]
[98]
[99]
[100]
[101]
[102]
[103]
[104]
[105]
[106]
[107]
[108]
[109]
[110]
[111]
[112]
absent pulmonary valve: evaluation with magnetic

resonance imaging. Tohoku J Exp Med 1997;182:
35 9.
Frank H, Salzer U, et al. Magnetic resonance imaging
of absent pulmonary valve syndrome. Pediatr Cardiol
1996;17:35 9.
Bruce CJ, Breen JF. Images in clinical medicine: aortic coarctation and bicuspid aortic valve. N Engl J
Med 2000;342:249.
Mahrholdt H, Wagner A, et al. Images in cardiovascular medicine. Magnetic resonance assessment of
cardiac function, infarct scar distribution, and ventricular remodeling in the setting of ischemic cardiomyopathy. Circulation 2003;107:e103 4.
Perin EC, Silva GV, et al. Assessing myocardial viability and infarct transmurality with left ventricular
electromechanical mapping in patients with stable
coronary artery disease: validation by delayed-enhancement magnetic resonance imaging. Circulation
2002;106:957 61.
Wagner A, Mahrholdt H, et al. MR imaging of myocardial perfusion and viability. Magn Reson Imaging
Clin N Am 2003;11:49 66.
Sandstede JJ. Assessment of myocardial viability by
MR imaging. Eur Radiol 2003;13:52 61.
Moran GR, Thornhill RE, et al. Myocardial viability
imaging using Gd-DTPA: physiological modeling of
infarcted myocardium, and impact on injection strategy and imaging time. Magn Reson Med 2002;48:
791 800.
Kumbasar B, Wu KC, et al. Left ventricular true
aneurysm: diagnosis of myocardial viability shown
on MR imaging. AJR Am J Roentgenol 2002;179:
472 4.
Kettles DI, Sheppard M, et al. Left ventricular aneurysm, aortic valve disease and coronary narrowing in
a patient with Hunters syndrome. Cardiovasc Pathol
2002;11:94 6.
Sorensen MB, Moat NE, et al. Images in cardiovascular medicine: false left ventricular aneurysm documented by magnetic resonance imaging. Circulation
2002;105:1734.
Altman CA, Ettedgui JA, et al. Noninvasive diagnostic features of partial absence of the pericardium. Am
J Cardiol 1989;63:1536 7.
Chartrand-Lefebvre C, Filion R, et al. Residents corner. Answer to case of the month #51. Congenital
complete absence of the left pericardium. Can Assoc
Radiol J 1997;48:353 6.
Faridah Y, Julsrud PR. Congenital absence of pericardium revisited. Int J Cardiovasc Imaging 2002;18:
67 73.
Gatzoulis MA, Munk MD, et al. Isolated congenital
absence of the pericardium: clinical presentation, diagnosis, and management. Ann Thorac Surg 2000;69:
1209 15.
Hauser M, Berger MF. Congenital absence of left
pericardium. J Belge Radiol 1997;80:296.
Iwafuchi Y, Okada Y, et al. A case of complete ab-
[113]
[114]
[115]
[116]
[117]
[118]
[119]
[120]
[121]
[122]
[123]
[124]
[125]
[126]
[127]
[128]
[129]
[130]
sence of the left pericardium coexisting with straight

back syndrome. Kokyu To Junkan 1991;39:1055 9.
Kornyei V, Kiss A, et al. Congenital absence of the
left pericardium: case report and review of the literature. Orv Hetil 1993;134:1703 7.
McIlhenny J, Campbell SE, et al. Pediatric case of the
day: congenital absence of the pericardium. AJR Am
J Roentgenol 1996;167:270, 274 5.
Raman SV, Daniels CJ, et al. Congenital absence of
the pericardium. Circulation 2001;104:1447 8.
Ratib O, Perloff JK, et al. Congenital complete
absence of the pericardium. Circulation 2001;103:
3154 5.
Victor AR, Osorio P, et al. Congenital absence of left
pericardium. Rev Port Cardiol 2003;22:801 10.
Moura C, Areias JC, et al. Pericardial effusion, arrhythmogenic right ventricular dysplasia and apical
diverticulum: an uncommon association. Cardiol
Young 2002;12:286 7.
De Biase L, Di Renzi P, et al. Cardiac magnetic resonance imaging diagnosis of a loculated pericardial
effusion in a patient with a postpericardiotomy syndrome. Ital Heart J 2002;3:435 6.
Neumann T, Chanbay A, et al. Paracardial lipodystrophy versus pericardial effusion in HIV positive
patients. Heart 2002;87:E4.
Chong HH, Plotnick GD. Pericardial effusion and
tamponade: evaluation, imaging modalities, and management. Compr Ther 1995;21:378 85.
Rokey R, Vick III GW, et al. Assessment of experimental pericardial effusion using nuclear magnetic
resonance imaging techniques. Am Heart J 1991;
121(4 Pt 1):1161 9.
Grainger RC. Congenital heart disease: general principles. In: Grainger RG, Allison DJ, editors. Diagnostic radiology, an Anglo-American textbook of
imaging. Edinburgh: Churchill Livingstone; 1986.
Baron MG. Plain film diagnosis of common cardiac
anomalies in the adult. Radiol Clin North Am 1999;
37:401 20.
Steiner RM, Gross GW, et al. Congenital heart disease in the adult patient: the value of plain film chest
radiology. J Thorac Imaging 1995;10:1 25.
Puvaneswary M, Leitch J, et al. MRI of partial anomalous pulmonary venous return (scimitar syndrome).
Australas Radiol 2003;47:92 3.
Petersen G, Martin U, et al. Intralobar sequestration
in the middle-aged and elderly adult: recognition and
radiographic evaluation. J Thorac Cardiovasc Surg
2003;126:2086 90.
Xu H, Jiang D, et al. Pulmonary sequestration:
three dimensional dynamic contrast-enhanced MR
angiography and MRI. J Tongji Med Univ 2001;21:
345 8.
Zhang M, Zhu J, et al. Contrast enhanced MR angiography in pulmonary sequestration. Chin Med J
(Engl) 2001;114:1326 8.
Ooi GC, Cheung CW, et al. Pulmonary sequestration:
diagnosis by magnetic resonance angiography and
[131]
[132]
[133]
[134]
[135]
[136]
[137]
[138]
[139]
[140]
[141]
[142]
[143]
[144]
[145]
[146]
[147]
539
computed tomography. Chin Med J (Engl) 1999;

112:668 70.
Lehnhardt S, Winterer JT, et al. Pulmonary sequestration: demonstration of blood supply with 2D and 3D
MR angiography. Eur J Radiol 2002;44:28 32.
Kamata S, Sawai T, et al. Extralobar pulmonary sequestration with venous drainage to the portal vein: a
case report. Pediatr Radiol 2000;30:492 4.
Kouchi K, Yoshida H, et al. Intralobar bronchopulmonary sequestration evaluated by contrast-enhanced
three-dimensional MR angiography. Pediatr Radiol
2000;30:774 5.
Ito F, Asaoka M, et al. Upper thoracic extralobar
pulmonary sequestration with anomalous blood supply from the subclavian artery. J Pediatr Surg 2003;
38:626 8.
Yukiiri K, Mizushige K, et al. Contrast transesophageal echocardiography in diagnosing congenital
enlargement of the right atrium: a case report. Angiology 2003;54:619 23.
Gayet C, Pillot M, et al. Diverticulum of the right
atrium: a case report and review of the literature. Arch
Mal Coeur Vaiss 1992;85:1479 82.
Nobel CK, Duerinckx AJ, et al. Dyspnea and chest
pain associated with lung mass. Chest 1998;114:
618 20.
Onuigbo WI. The spread of lung cancer to the heart,
pericardium and great vessels. Jpn Heart J 1974;15:
234 8.
Gualdi GF, Volpe A, et al. Computed tomography and
magnetic resonance in the TNM staging of pulmonary
carcinoma. Clin Ter 1992;141:493 8.
Kurzel RB, Parisi AF, et al. Echocardiography and
tumor invasion of the heart. JAMA 1976;236:2601.
Lestuzzi C, Viel E, et al. Left atrial invasion by lung
carcinoma through a pulmonary vein. Int J Cardiovasc Imaging 2001;17:107 10.
Testempassi E, Takeuchi H, et al. Cardiac metastasis
of colon adenocarcinoma diagnosed by magnetic resonance imaging. Acta Cardiol 1994;49:191 6.
Gindea AJ, Gentin B, et al. Unusual cardiac metastasis in hypernephroma: the complementary role of
echocardiography and magnetic resonance imaging.
Am Heart J 1988;116(5 Pt 1):1359 61.
Vetto JT, Heelan RT, et al. Malignant melanoma
metastatic to the right atrium: an asymptomatic solitary metastasis diagnosed incidentally by magnetic
resonance imaging. J Thorac Cardiovasc Surg 1992;
104:843 4.
Shih TT, Su CT, et al. Diagnosis of cardiac metastasis
by computed tomography: report of 5 cases. J Formos
Med Assoc 1990;89:392 8.
Rivkin A, Meara JG, et al. Squamous cell metastasis
from the tongue to the myocardium presenting as
pericardial effusion. Otolaryngol Head Neck Surg
1999;120:593 5.
Wilhelmi U, Fellner F, et al. Myxoid liposarcoma
with pericardial metastasis. Rontgenpraxis 1997;50:
119 21.
540
[148] Cutrone JA, Georgiou D, et al. Metastatic spread of

cervical carcinoma to the right ventricle and pulmonary arteries: diagnosis by ultrafast computed tomography. Am J Card Imaging 1995;9:275 9.
[149] Gomez-Jorge J, Mitchell SE. Embolization of a pulmonary artery pseudoaneurysm due to squamous cell
carcinoma of the lung. J Vasc Interv Radiol 1999;
10:1127 30.
[150] Grayet D, Ghaye B, et al. Systemic-to-pulmonary
venous shunt in superior vena cava obstruction revealed on dynamic helical CT. AJR Am J Roentgenol
2001;176:211 3.
[151] Katz ES, Shah A, et al. Bilateral pulmonary artery
compression and obstruction by tumor: diagnosis by
unusual Doppler flow patterns. J Am Soc Echocardiogr 2003;16:185 7.
[152] Stoblen F, Neumann K, et al. CT angiography of the
pulmonary artery in patients with bronchial carcinoma. Langenbecks Arch Surg 1997;114:1277 9.
[153] Giovagnoni A, Ercolani P, et al. Evaluation of the
pulmonary artery by cine MRI. J Comput Assist
Tomogr 1992;16:553 9.
[154] Hasegawa I, Eguchi K, et al. Pulmonary hilar lymph
nodes in lung cancer: assessment with 3D-dynamic
contrast-enhanced MR imaging. Eur J Radiol 2003;
45:129 34.
[155] Ohno Y, Adachi S, et al. Multiphase ECG-triggered
3D contrast-enhanced MR angiography: utility for
evaluation of hilar and mediastinal invasion of bronchogenic carcinoma. J Magn Reson Imaging 2001;13:
215 24.
[156] Edmondstone WM. Flitting radiographic shadows: an
unusual presentation of cancer in the lungs. Thorax
1998;53:906 8.
[157] Takahashi K, Furuse M, et al. Pulmonary vein and
left atrial invasion by lung cancer: assessment by
breath-hold gadolinium-enhanced three-dimensional
MR angiography. J Comput Assist Tomogr 2000;24:
557 61.
[158] Loscertales J, Jimenez-Merchan R, et al. Usefulness
of videothoracoscopic intrapericardial examination of
pulmonary vessels to identify resectable clinical T4
lung cancer. Ann Thorac Surg 2002;73:1563 6.
[159] Rofsky NM, Weinreb JC, et al. Aortic aneurysm and
dissection: normal MR imaging and CT findings after
surgical repair with the continuous-suture graft-inclusion technique. Radiology 1993;186:195 201.
[160] Prendergast BD, Boon NA, et al. Aortic dissection:
advances in imaging and endoluminal repair. Cardiovasc Intervent Radiol 2002;25:85 97.
[161] Sutherland D, Block PC. Images in clinical medicine:
aneurysm of a saphenous-vein bypass graft. N Engl J
Med 2001;344:1139.
[162] Bunce NH, Mohiaddin RH, et al. Contrast-enhanced
magnetic resonance angiogram of coronary artery bypass graft aneurysm. Circulation 2000;102:3148.
[163] Reddy GP, Steiner RM. Aneurysm of saphenous vein
coronary bypass graft: diagnosis by computed tomography. J Thorac Imaging 1999;14:147 9.
[164] Chan R, Common AA. Aneurysm of coronary artery

bypass graft presenting as a mediastinal mass: case
report. Can Assoc Radiol J 1998;49:181 4.
[165] Doyle MT, Spizarny DL, et al. Saphenous vein graft
aneurysm after coronary artery bypass surgery. AJR
Am J Roentgenol 1997;168:747 9.
[166] Ferreira AC, de Marchena E, et al. Saphenous vein
graft aneurysm presenting as a large mediastinal mass
compressing the right atrium. Am J Cardiol 1997;79:
706 7.
[167] Fattori R, Descovich B, et al. Composite graft replacement of the ascending aorta: leakage detection
with gadolinium-enhanced MR imaging. Radiology
1999;212:573 7.
[168] Nguyen V, Nguyen K. Plain film of intracardiac foreign bodies: the blurring effect. South Med J 1991;84:
651 3.
[169] Plewa MC, Stavros M, et al. Cervical prevertebral
soft-tissue measurements and chest radiographic findings in acute traumatic aortic injury. Am J Emerg
Med 1997;15:256 9.
[170] Golzarian J. Imaging after endovascular repair of abdominal aortic aneurysm. Abdom Imaging 2003;28:
236 43.
[171] Floris Vos AW, Wisselink W, et al. Cine MRI assessment of aortic aneurysm dynamics before and after
endovascular repair. J Endovasc Ther 2003;10:433 9.
[172] Fattori R, Napoli G, et al. Descending thoracic
aortic diseases: stent-graft repair. Radiology 2003;
229:176 83.
[173] Farner MC, Carpenter JP, et al. Early changes in abdominal aortic aneurysm diameter after endovascular
repair. J Vasc Interv Radiol 2003;14(2 Pt 1):205 10.
[174] Engellau L, Albrechtsson U, et al. Measurements
before endovascular repair of abdominal aortic aneurysms: MR imaging with MRA vs. angiography and
CT. Acta Radiol 2003;44:177 84.
[175] Vos AW, Wisselink W, et al. Aortic aneurysm pulsatile wall motion imaged by cine MRI: a tool to evaluate efficacy of endovascular aneurysm repair? Eur J
Vasc Endovasc Surg 2002;23:158 61.
[176] White CS. MR imaging of thoracic veins. Magn
Reson Imaging Clin N Am 2000;8:17 32.
[177] Fink C, Schmaehl A, et al. Images in cardiovascular
medicine. Pulmonary vein stenosis after radiofrequency ablation for atrial fibrillation: image findings
with multiphasic pulmonary magnetic resonance angiography. Circulation 2003;107:e129 30.
[178] Arentz T, Jander N, et al. Incidence of pulmonary
vein stenosis 2 years after radiofrequency catheter
ablation of refractory atrial fibrillation. Eur Heart J
2003;24:963 9.
[179] Kato R, Lickfett L, et al. Pulmonary vein anatomy in
patients undergoing catheter ablation of atrial fibrillation: lessons learned by use of magnetic resonance
imaging. Circulation 2003;107:2004 10.
[180] Dill T, Neumann T, et al. Pulmonary vein diameter
reduction after radiofrequency catheter ablation for
paroxysmal atrial fibrillation evaluated by contrast-
[181]
[182]
[183]
[184]
[185]
[186]
[187]
[188]
enhanced three-dimensional magnetic resonance imaging. Circulation 2003;107:845 50.

Tsao HM, Chen SA. Evaluation of pulmonary vein
stenosis after catheter ablation of atrial fibrillation.
Card Electrophysiol Rev 2002;6:397 400.
Yang M, Akbari H, et al. Identification of pulmonary
vein stenosis after radiofrequency ablation for atrial
fibrillation using MRI. J Comput Assist Tomogr
2001;25:34 5.
Baba Y, Miyazono N, et al. Altered flow dynamics of
intravascular contrast material to the liver in superior
vena cava syndrome: CT findings. Abdom Imaging
2000;25:146 50.
Yamada T, Takahashi K, et al. Focal hepatic hot
spot in superior vena cava obstruction: correlation
between radiocolloid hepatic SPECT and contrast-enhanced CT. Clin Nucl Med 1999;24:533 4.
Herbener TE, Basile V, et al. Abdominal case of the
day. Focal liver enhancement on contrast-enhanced CT
scan caused by obstruction of the superior vena cava
(SVC). AJR Am J Roentgenol 1997;169:250,253 4.
Bashist B, Parisi A, et al. Abdominal CT findings
when the superior vena cava, brachiocephalic vein,
or subclavian vein is obstructed. AJR Am J Roentgenol 1996;167:1457 63.
Haramati LB, White CS. MR imaging of lung cancer.
Magn Reson Imaging Clin N Am 2000;8:43 57, viii.
Boiselle PM. MR imaging of thoracic lymph nodes: a
[189]
[190]
[191]
[192]
[193]
[194]
541
comparison of computed tomography and positron

emission tomography. Magn Reson Imaging Clin N
Am 2000;8:33 41.
Pannu HK, Wang KP, et al. MR imaging of mediastinal lymph nodes: evaluation using a superparamagnetic contrast agent. J Magn Reson Imaging 2000;12:
899 904.
Takenaka D, Ohno Y, et al. Differentiation of metastatic versus non-metastatic mediastinal lymph nodes
in patients with non-small cell lung cancer using respiratory-triggered short inversion time inversion recovery (STIR) turbo spin-echo MR imaging. Eur J
Radiol 2002;44:216 24.
Erasmus JJ, McAdams HP, et al. MR imaging of
mediastinal masses. Magn Reson Imaging Clin N
Am 2000;8:59 89.
Knisely BL, Broderick LS, et al. MR imaging of the
pleura and chest wall. Magn Reson Imaging Clin N
Am 2000;8:125 41.
Pace ME, Krebs TL. MR imaging of the thoracoabdominal junction. Magn Reson Imaging Clin N Am
2000;8:143 62.
Mastora I, Remy-Jardin M, et al. Multi-detector row
spiral CT angiography of the thoracic outlet: dose
reduction with anatomically adapted online tube current modulation and preset dose savings. Radiology
2004;230:116 24.
Radiol Clin N Am 42 (2004) 543 564
Postoperative cardiopulmonary thoracic imaging

Anil Attili, AFRCS, FRCR, Ella A. Kazerooni, MD, MS*
University of Michigan, 1500 East Medical Center Drive, Ann Arbor, MI 48109 0326, USA
In recent years there has been tremendous growth

in the volume of cardiac surgery performed in the
United States [1]. Improvements in surgical and
anesthetic techniques have reduced the postoperative
mortality and morbidity associated with complex
cardiac surgical procedures. The radiologist plays a
significant role in the postoperative evaluation of
cardiac surgical patients. An appreciation of the
normal postoperative changes and complications following cardiac surgical procedures is essential when
interpreting postoperative imaging studies. This article focuses on both the normal postoperative appearances and the imaging of complications following
common cardiac surgical procedures. Most cardiac
surgery is currently performed by a median sternotomy using an extrapleural approach or rib thoracotomy, and cardiopulmonary bypass. Irrespective of the
specific nature of the surgery, certain common complications may occur, either from the surgery itself,
the use of cardiac bypass, or the patients underlying
cardiac disorder. Postoperative imaging after specific
thoracic cardiovascular surgeries is also discussed.
Early imaging findings after cardiac surgery

Immediately after cardiac surgery, patients are
admitted to a cardiac intensive care unit (ICU) where
the chest radiograph plays a key role in evaluation.
Comprehensive assessment of chest radiographs in
E-mail address: ellakaz@umich.edu (E.A. Kazerooni).
the cardiac ICU involves evaluation of indwelling

monitoring and therapeutic devices, fluid balance, the
pulmonary parenchyma, the cardiomediastinal silhouette, and extrapulmonary air and fluid collections.
Musculoskeletal structures and the visible upper
portions of the abdomen should also be evaluated.
Critical care pathways that include radiologic imaging have been implemented in cardiac ICUs for the
proper timing, sequencing, and use of resources. The
introduction of a critical care pathway has been
shown to decrease the use of bedside radiography
without adversely affecting short-term patient outcomes [2]. Little clinical value has been shown from
the use of routine chest radiographs after cardiac
surgery when performed routinely on admission to
the ICU and subsequently on a daily basis [3,4].
Chest radiographs are indicated for patients with
acute cardiopulmonary problems or on mechanical
ventilation, and after the insertion of new vascular
catheters and tubes, such as central venous and pulmonary artery catheters, intra-aortic balloon pumps
(IABP), and esophageal and endotracheal tubes [5].
A teamwork approach to patient care, with timely
and accurate interpretation of imaging studies,
and prompt communication with referring clinicians,
is essential.
Monitoring and therapeutic devices

These devices are critical to patient care in the
immediate postoperative period. The interpretation of
postoperative chest radiographs requires knowledge
of their normal position, and recognition of abnormal positioning and the complications that may
subsequently arise. Table 1 summarizes the lines and
doi:10.1016/j.rcl.2004.03.006
544
A. Attili, E.A. Kazerooni / Radiol Clin N Am 42 (2004) 543564
Table 1
Apparatus seen on postcardiac surgery chest radiographs
Location
Apparatus
Airway
Endotracheal tube
Tracheostomy tube
Central venous pressure catheter
Pulmonary artery catheter
Intra-aortic balloon pump
Extracorporeal life support cannulas
Temporary epicardial pacing leads
Left atrial catheter
Assist devices
Chest tubes
Drains
Drainage tubes
Feeding tubes
Venous
Arterial
Cardiac
Pleural
Mediastinum
Esophagus
Fig. 2. Endotracheal tube (arrow) with tip in the right main

bronchus. Note secondary hyperinflation of the right lung
with mediastinal shift to the left.
tubes that may be seen on chest radiographs in postcardiac surgery patients. Although some are commonplace, such as endotracheal tubes and central venous
catheters (Fig. 1), others, such as IABPs and ventricular assist devices, are less commonly encountered.
Airway
Endotracheal tubes provide respiratory support
and airway protection in the immediate postoperative
period. The normal position of an endotracheal tube
is with the tip 2 to 6 cm above the carina. Complications of endotracheal tube placement are as follows:
Vocal cord injury
Right main bronchus intubation (Fig. 2)
Contralateral lung collapse
Ipsilateral pneumothorax
Esophageal intubation
Airway rupture
Delayed airway stenosis
In the past, extubation usually was deferred until
the day after cardiac surgery. Today, patients frequently are weaned from the ventilator and extubated
soon after leaving the operating room [6]. Most
low- and moderate-risk patients undergoing cardiac
surgery with cardiopulmonary bypass using opioid
analgesia are extubated within 7 to 11 hours after
operation. This has been shown to reduce the length
of stay in the ICU with no increase in postoperative
complications [7]. Prolonged ventilatory support in
the postoperative period may necessitate a tracheostomy. Complications of tracheostomy tubes include
positioning in soft tissues outside the airway, airway
stenosis, and trachea-innominate artery fistula.
Vascular
Venous
Fig. 1. Typical apparatus seen on post cardiac surgery
radiographs. Note the normal position of the endotracheal
tube (ET), Swan-Ganz catheter (SG), mediastinal drains
(M), and esophageal tube entering the stomach (T).
Central venous catheters are essential for monitoring, modifying fluid and pressure balances, and
administering therapeutic substances, such as drugs
and antibiotics. The tip of a central venous catheter

should be located in the superior vena cava or a
brachiocephalic vein. Complications of central venous catheters are as follows [8]:
Pneumothorax
Hemothorax
Cardiac perforation and tamponade
Pneumomediastinum and hemomediastinum
Mural thrombus
Venous dissection
Cardiac arrhythmias
Catheter fracture, knotting, and embolization
Infection
Pulmonary artery catheters provide valuable information on cardiac function and hemodynamic
status in the postoperative period by measuring
pulmonary artery pressure, pulmonary resistance,
cardiac output, and pulmonary capillary wedge pressure. The latter is an indicator of left ventricular
end-diastolic pressure and left ventricular function.
Ideally, the tip of a pulmonary artery catheter should
reside within a large pulmonary artery and should not
be located peripheral to the interlobar pulmonary
artery that forms the hilum of each lung. The balloon
should only be inflated during placement and pressure measurement, and should not be left wedged
in a small artery where arterial wall damage may occur. In addition to the complications of central venous
catheters, complications unique to flow-directed pulmonary artery catheters (Figs. 3 and 4) include
pulmonary artery perforation with hemorrhage,
pulmonary infarcts, and pulmonary artery pseudoaneurysms. The latter are treated with percutaneous
catheter-guided embolization techniques to reduce the
risk of rupture and thrombus formation [9].
Fig. 3. Swan-Ganz catheter coiled across the tricuspid valve

with tip (arrowhead) in the right pulmonary artery.
Redundancy may be thrombogenic and impair valve
function. The catheter was subsequently replaced.
545
Left atrial catheters are used uncommonly for

direct measurement of left atrial pressure. They may
be used when there is difficulty traversing the right
heart with a pulmonary catheter, such as after tricuspid valve repair [10]. The catheter is usually
inserted into the left atrium at the time of surgery,
either through needle puncture of the right superior
pulmonary vein or through a direct puncture of the
left atrium. The external end of the catheter is brought
out through the chest wall or the inferior end of the
sternal incision. These devices can typically be visualized on chest radiographs as thin, curvilinear metallic lines within the left atrium. Complications
include line fragmentation and retention, embolization, infection, and cardiac tamponade [11].
Cardiac
Temporary epicardial pacing wires are frequently
placed during cardiac surgery to diagnose and treat
arrhythmias, and to enhance cardiac output in the
postoperative period [12]. They are seen adjacent to
the right atrium and right ventricle on postoperative
radiographs. Complications include bleeding, pneumothorax, and pneumomediastinum. Although they
are usually removed before discharge, occasionally
they do not come out easily when pressure is placed
on the external port, in which case they are cut off
at the skin surface and the internal portion left in
place. These retained wires are a relative contraindication to future MR imaging examinations and
should be noted in the radiographic report [13].
Pump failure in the postoperative period despite
maximal inotropic support is an indication for mechanical circulatory support, such as an intraaortic
balloon pumps (IABP) or a left ventricular assist
device (LVAD) [14]. IABPs are the most widely
used form of mechanical circulatory support in cardiac surgery patients during the perioperative period,
currently used in 2% to 12% of all patients undergoing cardiac surgery [15]. Preoperative insertion may
be performed in high-risk patients. An IABP uses
the principle of diastolic counterpulsation in which
the balloon inflates in synchrony and out of phase
with the cardiac cycle. Benefits of this technique
include augmentation of diastolic coronary perfusion
pressure, reduced systolic afterload, and increased
cardiac output with an improved myocardial oxygen
supply:demand ratio. IABPs are usually inserted
percutaneously through the common femoral artery
using a 9.5F catheter. In patients with small vessels,
an 8.5F catheter is used. A chest radiograph is
obtained following insertion to ensure correct positioning. The radiopaque tip of the IABP should be
546
Fig. 4. Pulmonary artery pseudoaneurysm in a 71-year-old man with hemoptysis after Swan-Ganz catheter placement.
(A) Radiograph showing catheter in right pulmonary artery (arrow). The patient subsequently moved in bed and had an episode
of hemoptysis. The catheter was found wedged and subsequently pulled back. (B) Radiograph 2 hours later demonstrates new
opacification of the right upper lobe caused by hemorrhage; the Swan-Ganz catheter is in the pulmonary outflow tract (arrow).
(C) Pulmonary angiogram demonstrates a pseudoaneurysm of the anterior branch of the right upper lobe pulmonary artery.
(D) Radiograph 48 hours after embolization demonstrates resolving pulmonary hemorrhage and embolization coils.
located in the proximal descending thoracic aorta

distal to the left subclavian artery and at the inferior
aspect of the aortic knob (Fig. 5). It should project
over the aortic knob on frontal radiographs [16].
Complications of IABPs are listed as follows:
Leg ischemia
Cerebral emboli and stroke
Mesenteric and renal ischemia
Aortic rupture, dissection, pseudoaneurysm
Balloon rupture
Septicemia
Reported complication rates of IABPs vary from
12.9% to 29% [17,18]. Leg ischemia is by far the
most common complication, with an incidence of

9% to 25%. Most radiographically apparent complications result from improper positioning of the balloon tip [19]. Positioning too high may result in
extension into the aortic arch branches, potentially
injuring or occluding the left subclavian or left
vertebral arteries, resulting in embolization and
stroke. Too distal a position may result in bowel
ischemia or renal embolism because of arterial occlusion by the balloon. Loss of definition of the
descending thoracic aorta on the chest radiograph
should raise suspicion of intramural positioning,
which may lead to aortic dissection or rupture.
The IABP is an attractive form of circulatory
support because of its ease of insertion and removal.
Fig. 5. Intra-aortic balloon pump and extracorporeal life

support (ECLS) in a 44-year-old woman with biventricular
dysfunction 4 hours after orthotropic cardiac transplantation for nonischemic cardiomyopathy. Note position of
the radiopaque tip of the IABP at the inferior aspect of
the aortic knob (arrowhead), and ECLS cannulae placed
during sternotomy in the right atrium (short arrow) and
aorta (long arrow).
However, it yields only a modest increase in cardiac output. Failure of the IABP to improve hemodynamic performance of the failing heart may prompt
use of an alternative form of mechanical circulatory
support, such as a centrifugal pump, extracorporeal
life support, pneumatic pulsatile pumps, or an implantable LVAD.
Centrifugal pumps are the second most commonly
used cardiac-assist devices. Centrifugal pumps impart
momentum to fluid by means of blades, impellers, or
concentric cones, yielding continuous nonpulsatile
flow. The left ventricular support system consists of
a cannula that siphons blood from the heart, usually
from the left atrium; a pump that drives the blood
back into the arterial circulation under pressure; and a
return cannula connected to the aorta or femoral
artery [20]. When both ventricles require support, a
separate cannula siphons the blood from the right
atrium and passes it to a second pump. The blood is
then returned under pressure to the pulmonary artery.
Extracorporeal membrane oxygenators for extracorporeal life support provide combined heart and lung
support, with similar hospital survival rates to centrifugal mechanical support in the setting of postcardiotomy cardiogenic shock. They can be used to
sustain patients until a long-term form of LVAD can
be placed [21]. Venovenous extracorporeal life sup-
547
port is used for respiratory support alone. Venous

blood is siphoned from the right atrium through a
cannula placed through the right internal jugular
vein. After oxygenation blood is returned to a large
peripheral vein, usually a femoral vein, or directly to
the right atrium through a double-channel catheter
that is also used to remove the blood. Venoarterial
extracorporeal life support is used to support both the
cardiovascular and respiratory systems. Blood is
removed from the right atrium, oxygenated, and
pumped back into the ascending aorta under arterial
pressure through a large-bore catheter usually placed
in the right carotid artery. On a chest radiograph the
normal location of the venous cannula tip should be
in the distal superior vena cava or the right atrium
(Fig. 6). The aortic cannula should be at the top of the
aortic arch or in the innominate artery immediately
adjacent to the aortic arch [22]. There may be a
nonradiopaque tip on the cannulae, and they may
extend further than is radiographically apparent. The
thoracic complications of extracorporeal life support
are usually related to bleeding, such as hemothorax,
because extracorporeal life support requires systemic
anticoagulation [23]. During extracorporeal life support the lungs become diffusely opacified because of
atelectasis and a systemic response, unrelated to any
underlying lung disease.
Fig. 6. Adult respiratory distress syndrome after coronary

artery bypass graft surgery, requiring venovenous extracorporeal life support in a 48-year-old male patient. The tip
of ECLS cannula is in the distal superior vena cava (arrow).
Also note Swan-Ganz catheter in the right pulmonary artery, left chest wall defibrillator, tracheostomy, and esophageal tubes.
548
Pneumatic pulsatile pumps include the Thoratec

system (Thoratec Laboratories, Pleasanton, California) and the Abiomed system (Abiomed, Danvers,
Massachusetts) (Fig. 7). Radiographs of patients with
these pumps are indistinguishable from centrifugal
pumps. The drainage and delivery cannulae are similarly placed, with the pump apparatus outside the
field of view [20]. Implantable assist devices allow
mobility and are of two basic types: electromechanical (Novacor LVAD, World Heart, Ottawa, Ontario,
Canada) and pneumatic (Heartmate LVAD, Thoratec,
Pleasanton, California) (Fig. 8). Both types have
similar blood flow systems outside of the pumping
chamber itself. Blood comes to the pump through a
conduit exiting from the left ventricular apex. The
blood is pumped back into the circulation through
a cannula placed in the ascending aorta or less
commonly the abdominal aorta. Recent results with
the implantable LVAD systems in postcardiotomy
shock applied within the context of a well-designed,
highly interactive postcardiotomy network have been
promising [24]. The Debakey Micromed LVAD
(Micromed Technology, Houston, Texas) is a miniature novel implantable LVAD with the potential to
address the limitations of the larger pulsatile pumps
(Fig. 9). Multicenter clinical trials of the Debakey
Micromed LVAD as a bridge to cardiac transplantation are underway and initial results have shown it to
be successful with a low complication rate [25].
Fig. 7. Abiomed left ventricular assist device (LVAD)

(Abiomed, Danvers, MA) and venovenous extracorporeal
life support in a 37-year-old woman with dilated cardiomyopathy. The LVAD pump apparatus is outside the field
of view. Note the radiopaque cannula (arrow) entering the
left atrium through the right superior pulmonary vein. The
return cannula is nonradiopaque and returns the blood to the
ascending aorta. The tip of the ECMO cannula is in the right
atrium (arrowhead).
Fig. 8. HeartMate LVAD (Thoratec, Pleasanton, CA) in a

31-year-old woman with idiopathic cardiomyopathy complicated by cardiogenic shock. She subsequently underwent cardiac transplantation. CT scout image demonstrates
the pumping chamber and opaque proximal portions of the
afferent (A) and efferent cannulae (E).
Fig. 9. DeBakey micromed LVAD (Micromed Technology,

Houston, TX) in a 63-year-old woman with ischemic cardiomyopathy and cardiogenic shock. The pump apparatus (P) and afferent cannula (A) of the DeBakey micromed
LVAD are visible. The efferent cannula is nonradiopaque
and enters the ascending aorta. The patient subsequently
underwent a cardiac transplant.
Mediastinal and pleural catheters

Two or three mediastinal tubes with radiopaque
stripes are commonly inserted into the mediastinum
at the completion of surgery, just before sternal closure (see Fig. 1). One or two tubes drain the anterior
mediastinum and are directed superiorly near the
midline. A posterior right-angled drainage tube also
may be placed between the diaphragm and inferior
heart border to drain the inferior and posterolateral
pericardium. A pleural drainage tube may be present
in either pleural space if a pneumothorax was entered
during surgery, which may occur during internal
mammary artery grafting or sternal reapproximation.
Both mediastinal and pleural tubes enter inferior to
the midline incision when introduced at surgery.
Esophageal tubes
An esophageal tube is usually seen coursing
through the thorax into the left upper quadrant. The
tip should be located in the stomach for gastric
drainage. Feeding tubes are not used after uncomplicated surgery, in which patients are frequently
transferred from the ICU to a regular hospital bed in
48 to 72 hours. Complications of esophageal tubes
include malpositioning within the gastrointestinal
tract, positioning within the airway, lung laceration or
hemorrhage, pneumothorax, hemothorax, and esophageal rupture.
Coronary artery bypass graft markers
The location of the vein graft anastomoses on the
aorta may be marked with either large wire circles
around the ostia (see Fig. 16) or small washer-like
markers. Such markers are useful to the cardiologists
when localizing the grafts for catheterization during
subsequent coronary angiography. Patients with such
markers have been shown to require significantly less
fluoroscopy time and contrast volume when undergoing cardiac catheterization compared with patients
without markers [26]. Hemostatic clips used to ligate
the intercostal arteries when the internal mammary
artery is used for bypass grafting are commonly seen
on radiographs. This line of clips is visible on postoperative chest radiographs and usually bows convex
to the right along the left side of the mediastinum.
Pulmonary parenchymal opacification
Fluid balance in the postoperative period can be
estimated by the degree of vascular distinctness,
peribronchial cuffing, presence or absence of inter-
549
stitial or alveolar edema, the vascular pedicle width,

and the thickness of the chest wall [27]. On immediate postoperative radiographs, almost all patients
exhibit mild interstitial pulmonary edema caused by
increased capillary leak following cardiopulmonary
bypass. The lungs clear with improving vascular
distinctness over the next 1 to 2 days, provided that
the surgical repair was effective, fluid balance is wellmaintained, and the patient has adequate renal function. Patients with pump failure in the postoperative
period exhibit overt interstitial and alveolar edema
radiographically. Delayed pulmonary edema may
develop several days following cardiac surgery, particularly in elderly patients. This develops because of
relative hypovolemia after surgery caused by blood
loss, insensible loss at surgery, the cardiopulmonary
bypass itself, and capillary injury. Fluid is generally
required clinically, a portion of which ends up in the
extravascular space. After several days this fluid is
resorbed into the vascular space, resulting in hypervolemia, hypoalbuminemia, and pulmonary edema.
In patients undergoing cardiac valve surgery, fluid
balance is exceptionally difficult because of the
unpredictability of response to the surgical repair. In
general, stenotic lesions respond to repair more
predictably than mixed or regurgitant lesions [28].
Acute respiratory distress syndrome
After cardiac surgery, acute respiratory distress
syndrome has an incidence of 0.4% to 2.5%, with a
mortality of 15% to 34% [29,30]. It manifests the
same way radiographically as other patients with
acute respiratory distress syndrome, as diffuse bilateral alveolar opacities (see Fig. 6). Predictors of acute
respiratory distress syndrome following cardiac surgery include redo surgery, poor preoperative respiratory function, blood transfusions, shock, smoking,
diabetes, renal failure, and poor left ventricular function [29].
Atelectasis
Postoperative pulmonary opacities caused by atelectasis are common in the left lower lobe, occurring
in 75% of patients. Another 10% to 20% of patients
develop bibasilar atelectasis [30]. The atelectasis
usually clears in a few days, but occasionally takes
several weeks. Postoperative atelectasis can be minimized if the patient is on positive end-expiratory
pressure. Major lobar collapse in areas other than the
left lower lobe should suggest mucous plugging. True
postoperative left lower lobe pneumonia is uncommon. Lordotic angulation of the X-ray beam during
550
portable radiography may create an illusionary opacity in the left retrocardiac region that can be misinterpreted as atelectasis or consolidation behind the
heart [31].
Pneumonia
Nosocomial pneumonia has an incidence of approximately 4% in patients undergoing coronary
bypass graft surgery [32]. Radiographic confirmation is sometimes difficult because of the frequent
coexistence of atelectasis and edema. Clinical correlation and comparing changes on sequential radiographs are helpful.
Pulmonary embolism
Pulmonary embolism occurs infrequently after
cardiac surgery, with an incidence 0.56%. It carries
a high mortality of 34% [33]. Risk factors include
preoperative bed rest, recent cardiac catheterization,
and postoperative congestive heart failure. The relatively low incidence of pulmonary embolism after
cardiac surgery has been attributed to intraoperative
heparinization and postoperative anticoagulation or
antiplatelet therapy. In addition, cardiopulmonary bypass causes changes in blood elements that may retard
clotting. These changes include consumption of coagulation factors, activation of the fibrinolytic cascade, and thrombocytopenia and platelet dysfunction.
The cardiomediastinal silhouette
The cardiomediastinal silhouette is an important
postoperative guide to the general well-being of the
patient. A change in its size or shape may indicate
mediastinal hemorrhage or cardiac tamponade. After
surgery the mediastinum appears slightly wider than
on the preoperative radiograph, in part because of the
anteroposterior supine technique used for portable
radiography. In addition, some mediastinal bleeding
normally occurs. The mediastinal width may be
reduced if the patient is on positive end-expiratory
pressure, and may increase slightly following extubation if the lung volumes decrease. Katzberg et al
[30] related the postoperative mediastinal width to the
severity of bleeding by comparing the width of the
mediastinum on preoperative posteroanterior radiographs with postoperative anteroposterior radiographs. Stable patients without clinical evidence of
bleeding widen their mediastinum by an average
of 35%. Patients with moderate bleeding of 30 to
280 mL who did not require reoperation had an
average 47% increase in width, and patients requiring
re-exploration had an average 60% increase in width.

All patients with over 70% increase in width required
reoperation. In some patients considerable blood loss
occurred with only mild to moderate radiographic
changes, whereas other patients with marked widening did not require re-exploration. Comparison of
preoperative and postoperative radiographs is difficult. The initial postoperative film is the best baseline
for subsequent changes in mediastinal width. The
mediastinum may not widen if hemorrhage decompresses into the pleural space or the extrapleural soft
tissues. Additional signs of mediastinal hemorrhage
include an increase in apical soft tissue opacity or
pleural fluid. The decision to reoperate is ultimately
based on the overall clinical picture and not mediastinal width alone.
Pericardial tamponade has been reported to occur
in 3.5% of patients undergoing cardiac surgery. It
most commonly occurs in the immediate postoperative period, presenting as an acute surgical emergency
[34]. Removal of temporary pacing wires or left atrial
lines may lead to cardiac tamponade 1 to 2 days after
cardiac surgery. Patients with early tamponade usually have excessive blood loss through indwelling
tubes and rising central venous pressure together with
hypotension, tachycardia, decreasing urine output,
and acidosis. In almost half of these patients there
is no gross alteration in cardiomediastinal width,
because a small amount of pericardial fluid can cause
acute tamponade without an apparent increase in
heart size [35]. Single-chamber cardiac tamponade
also is a complication of cardiac surgery [36]. The
chest radiograph in isolated right atrial tamponade
may demonstrate enlargement or increased convexity
of the right heart border when compared with preoperative radiographs. Echocardiography to evaluate
for pericardial fluid may be technically difficult immediately following surgery because of mediastinal
drains and pneumomediastinum. CT is useful to demonstrate pericardial fluid in this circumstance and
to detect focal intrapericardial hematoma causing
single-chamber compression [37].
Traumatic venous catheter insertion may produce
a hematoma with alterations in mediastinal contour.
Nonhemorrhagic events may also widen the mediastinum, such as aortic dissection complicating aortic
valve repair, at the site of proximal insertion of a
coronary vein graft, aortic bypass cannula, or from
the site of aortic cross-clamping.
Cardiac herniation through a pericardial defect is a
rare catastrophic complication that usually occurs in
the first few hours following cardiac surgery [35].
Abnormally sharp notches between the herniated heart
and the mediastinum, and shift of the heart toward the
herniated side, may be seen on radiographs. Prompt

corrective surgery is usually lifesaving.
Extrapulmonary fluid and air collections
A small pleural effusion is very common in the
postoperative period, particularly on the left side. It
usually regresses within several days. Pneumothorax,
pneumomediastinum, and subcutaneous emphysema
are frequently present following median sternotomy.
These collections are usually self-limiting and resolve
over a few days. Air has been reported to persist for
50 days, however, in the retrosternal soft tissues
without clinical significance [38]. Soft tissue air
collections that appear de novo or progressively
increase should raise concern for infection with gasforming organisms. Expanding pneumothorax or interstitial pulmonary emphysema may be the result of
barotrauma in patients requiring high levels of positive end-expiratory pressure or high peak inspiratory
pressure during mechanical ventilation. A benign
self-limiting pneumoperitoneum may occasionally
occur after cardiac surgery as a result of a long
incision extending below the diaphragm or after
subxiphoid tunneling for epicardial pacemaker [39].
Musculoskeletal structures
Radiographically occult rib fractures are common
after median sternotomy. In one study of 24 patients
undergoing bone scans following median sternotomy,
551
44 total rib fractures were demonstrated, with 30 on

the left side and 14 on the right side. The upper three
ribs were the most often involved. In retrospect,
only four of these fractures were visible on chest
radiographs [40]. The presence of an extrapleural
hematoma may be a telltale sign of an underlying
rib fracture. Occult rib fractures may be a cause of
postoperative nonincisional pain. Other serious causes
of nonincisional pain, such as pericarditis, myocardial infarction, and pulmonary embolism, must first
be excluded.
Delayed complications
Sternal dehiscence, osteomyelitis, and mediastinitis
Sternal dehiscence, osteomyelitis, and mediastinitis are interrelated but uncommon serious postoperative complications. They are associated with a high
mortality and morbidity [38]. The diagnosis of sternal
dehiscence may be evident on physical examination.
The two major radiographic signs of dehiscence are
the mid-sternal stripe sign [41] and sternal wire
displacement [42]. The latter is highly specific
(Fig. 10). A recent study on the frequency of sternal
wire abnormalities in patients with sternal dehiscence concluded that sternal wire abnormalities, most
notably displacement, are present in most patients
with sternal dehiscence and that radiographic abnormalities precede the clinical diagnosis in most cases
Fig. 10. Sternal dehiscence with sternal wire displacement on the frontal chest radiograph. (A) Midline vertically aligned
sternotomy wires after cardiac surgery. (B) Later, there is malalignment of the upper two sternal wires (arrows).
552
Pericardial complications
Fig. 11. Mediastinitis and an infected aortic graft in a

71-year-old woman with fever and chills 6 months after
ascending aorta and aortic root replacement. Axial contrast-enhanced CT demonstrates low-attenuation material
(arrowheads) and several air bubbles (arrows) surrounding
the aortic graft, strongly suggestive of graft infection
and mediastinitis.
[42]. A mid-sternal lucent stripe thicker than 3 mm

should raise suspicion for sternal dehiscence. This
sign is uncommon in dehiscence, however, and does
not add incremental value to the finding of sternal
wire displacement alone [41].
The reported incidence of mediastinitis following
cardiac surgery is 0.4% to 5%, with a mortality of 7%
to 80% [38]. CT plays a major role in the assessment
of mediastinitis by depicting the extent and depth of
abnormality. The anterior mediastinum is universally
abnormal after median sternotomy. Focal retrosternal
fluid collections, pneumomediastinum, hematoma, or
any combination of these three may be found up to 21
days after surgery in asymptomatic individuals [43].
The specificity of CT is time dependent; it is important to integrate the CT findings with the clinical
course. The CT findings of mediastinitis (Fig. 11)
include obliteration of mediastinal fat planes, lowattenuation mediastinal fluid collections, air and fluid
collections (mediastinal abscess), and sternal separation [43,44]. CT also may be used to guide percutaneous aspiration and drainage.
Sternal osteomyelitis may be an isolated problem
or may be associated with sternal dehiscence or
mediastinitis. Early sternal osteomyelitis on CT can
be difficult to differentiate from minor sternal irregularities caused by the sternal osteotomy and normal
anatomic variants. Eventually, frank bone destruction,
severe demineralization, and dehiscence occur [44].
When CT findings are equivocal, gallium 67 scanning may be useful [45].
Postpericardiotomy syndrome, pericardial effusions, tamponade, and constrictive pericarditis are

the late pericardial complications of cardiac surgery.
Postpericardiotomy syndrome has an incidence of
10% to 40% after cardiac surgery [46]. It is characterized by fever, pericarditis, pleural effusion, and
pleuritis. Although it usually develops 2 to 3 weeks
after cardiac surgery, uncommonly it may develop up
to 6 months later [35]. Pericardial effusion, pleural
effusion, and occasionally basilar pulmonary opacities are seen on radiographs. These findings are
nonspecific, and radiographs may also be normal.
Treatment is usually nonsteroidal anti-inflammatory
agents, with corticosteroids reserved for persistent
effusions. Although it is usually self-limited, hemodynamically significant pericardial effusions may
develop, leading to cardiac tamponade [47].
The incidence of late hemodynamically significant pericardial effusions, developing more than
7 days after cardiac surgery (Fig. 12), ranges from
0.1% to 6% [48]. Anticoagulant use is an important
contributing factor in the approximately two thirds of
these cases. The postpericardiotomy syndrome causes
a large number of these late effusions, and accounts
for one third of effusions that present after the first
postoperative week [47]. Echocardiography can
quickly confirm the presence of an effusion. Pericardiocentesis under echocardiographic guidance is successful in most cases. The use of a pericardial catheter
for extended drainage is associated with a lower
Fig. 12. Pericardial effusion after cardiac surgery in a

35-year-old man 10 weeks after atrial septal defect repair with
a large pericardial effusion and a small right pleural effusion.
553
Fig. 13. Loculated pericardial effusion (*) compressing the

right ventricle on CT in a 62-year-old man caused by
purulent pericarditis.
recurrence rate [47]. The pericardial effusion may be

global or it may be localized, manifesting as a
localized deformity radiographically [36]. CT is useful in such cases (Fig. 13) to define the location and
extent of hematoma [37].
Constrictive pericarditis is a late complication of
cardiac surgery, with an incidence of 0.2% to 2%
[49]. It develops an average of 82 days after surgery.
Chest radiographs are not helpful in diagnosis, unless
pericardial calcification is identified. Pericardial
thickening is the hallmark of constrictive pericarditis
on CT (Fig. 14) and MR imaging, defined as pericardial thickness of 4 mm or greater. MR imaging
provides functional information and allows differen-
Fig. 15. Multidetector coronary CT angiogram reconstruction demonstrates a patent right internal mammary artery
coronary bypass graft with adjacent surgical clips.
tiation between constrictive pericarditis and restrictive cardiomyopathy [50].

Aortic pseudoaneurysm and dissection
An incision in the aortic wall during cardiac
surgery predisposes to aortic dissection or pseudoaneurysm formation. The incision may be an aortotomy during aortic valve replacement or cannulation
for cardiopulmonary bypass. The incidence of aortic
Fig. 14. (A,B) CT of constrictive pericarditis secondary to prior cardiac surgery in a 61-year-old man who developed right heart
failure 3 months after replacement of the aortic root and ascending aorta for type A dissection. Axial contrast-enhanced CT
images demonstrate pericardial thickening (arrowheads) and a small amount of pericardial fluid (F). Note the dissection in the
descending aorta and small bilateral pleural effusions.
554
dissection after cardiac surgery is approximately

0.8%; pseudoaneurysms following cardiac surgery
are less common [51]. Mediastinal infection is a
risk factor for pseudoaneurysm formation. They
may present weeks or years after cardiac surgery.
Chest radiographs may show focal or diffuse mediastinal widening. Intravenous contrast-enhanced CT
is the procedure of choice for diagnosis and evaluation [52].
Postoperative imaging after specific cardiac

surgeries
Coronary artery bypass graft surgery
Catheter angiography is the gold standard for
detecting progressive atherosclerosis and to evaluate
graft patency after coronary artery bypass. Retrospective electrocardiographic-gated multidetector spiral
CT permits the noninvasive assessment of bypass
graft patency and stenosis with high diagnostic accuracy (Fig. 15). This method is still limited, however,
by a significant number of bypass grafts that are
unevaluable for the presence or absence of significant
stenosis [53]. With improvements in surgical techniques and medical care, an increasing number of
patients are now candidates for reoperative cardiac
surgery. Defining the anatomy of pre-existing grafts
is critical in the reoperative cardiac surgery patient,
because injury to these vital structures is associated
with significant postoperative morbidity and mortality. Internal mammary grafts are at particular risk;
however, accurate evaluation of saphenous vein
grafts is equally important. The anatomic relationship
of the grafts to the sternum must be assessed accurately to prevent injury during sternal reentry. The use
Fig. 16. Saphenous vein graft pseudoaneurysm in a 69-year-old man at the proximal anastomosis of a graft to the right coronary
artery. (A) Posteroanterior and (B) lateral chest radiographs demonstrate an anterior mediastinal mass (*). Note the ring-shaped
proximal graft markers. (C) Contrast-enhanced CT demonstrates an anterior mediastinal mass with central contrast enhancement
(arrow) that is contiguous with the coronary graft ostium extending from the ascending aorta. Note the mural thrombus (T).
of noninvasive electrocardiographically gated multidetector CT with volume-rendered images allows

complete anatomic mapping of the grafts, providing
valuable information to the operating surgeon [54].
Saphenous vein graft aneurysms are a well-recognized complication after bypass surgery. They have
been reported from 4 months to 21 years after surgery
[55]. The clinical presentation is often that of myocardial ischemia with angina. Many graft aneurysms
are asymptomatic, however, and detected incidentally
on chest radiographs, CT, or angiography. On chest
radiographs, a graft aneurysm may appear as a paracardiac, hilar, or mediastinal mass (Fig. 16). CT
demonstrates an enhancing, round mass along
the heart border that may either uniformly enhance
with intravenous contrast or be of heterogeneous attenuation because of internal thrombus (Fig. 10C).
Coronary angiography confirms the diagnosis. Mural
thrombus, however, may obscure the true dimensions
of the aneurysm at angiography. Pseudoaneurysms at
the proximal and distal ends of grafts are more
common than true aneurysms involving the body of
the graft. The treatment options include resection and
revascularization, thrombectomy, and therapeutic embolization [56].
Aortic reconstruction
Surgical repair of aortic dissections and aneurysms has substantially improved the course of these
Fig. 17. Pseudoaneurysm at the distal anastomosis of an

aortic graft in a 57-year-old man status post descending
thoracic aorta replacement for an aneurysm. The patient
presented 20 years later with increasing chest discomfort.
Oblique sagittal reconstruction from a contrast-enhanced multidetector CT demonstrates the pseudoaneurysm (asterisk).
555
Fig. 18. High-attenuation felt (arrowheads) at an aortic graft

anastomosis in the ascending aorta is a normal appearance.
life-threatening conditions. Imaging the postoperative

aorta enables monitoring of aortic diameters and early
detection of complications. Accurate postoperative
imaging evaluation requires knowledge of the surgical technique used and the anatomic consequences.
Portions of the aorta may be resected or opened,
grafts may be sewn end-to-end or end-to-side, and
branch vessels may be reimplanted or grafted [57,58].
Inclusion techniques involve aortotomy, graft insertion, and subsequent enclosure of the graft by the
remnant of the diseased aorta, resulting in a potential
space between the graft and the aortic wall. Graft
interposition techniques involve total excision of a
segment of native aorta with graft anastomosis to the
proximal and distal excision sites without a concomitant wrap.
The CT and MR imaging appearances of the postoperative aorta following repair using the continuous-suture graft inclusion technique have been well
described [59]. The potential space between the graft
and the native aorta may contain a small amount of
blood and gas in the immediate postoperative period.
The presence of gas in this space more than 2 weeks
after surgery is virtually pathognomonic of infection
[59]. Pseudoaneurysm formation is a major complication of graft-interposition technique, resulting from
partial dehiscence of a suture line (Fig. 17). Perigraft
blood flow manifesting as intravenous contrast outside the expected confines of the graft lumen and
perigraft thickening are other observed findings [59].
The synthetic interposition grafts of the aorta have
a characteristic appearance on CT [57]. Felt pledgets
556
and strips used to reinforce sutures appear as highattenuation material bordering the wall of the aorta or
the graft (Fig. 18), and should not be confused with
extravascular contrast material from a leaking graft
(Fig. 19). Kinking of the graft or puckering of the
anastomosis may create the appearance of a transverse low-attenuation band traversing the aorta on
axial images, mimicking dissection. Multiplanar CT
reconstructions are useful to avoid making this mistake. The button technique, where a small portion of
the native aorta around the coronary artery ostium is
implanted onto the graft, may create the appearance
of an outpouching from the graft on CT or MR
imaging. Failure to recognize the relationship of the
outpouching to the coronary artery may result in the
misdiagnosis of a pseudoaneurysm. Circumferential
low-attenuation or soft tissue material surrounding or
adjacent to the graft on CT may be seen for months to
years following surgery and should not be mistaken
for leak or infection. Other mimics of pathology
include the collapsed native aorta adjacent to a graft
(Fig. 20), and reinforcement of the graft with bovine
pericardium (Fig. 21). Endovascular stent management of thoracic aortic aneurysms and dissections
(Fig. 22) may be used in patients with multiple
comorbidities as an alternative to graft placement,
to reduce the incidence of negative surgical outcomes
in these high-risk patients.
Fig. 19. Rupture of an aortic graft in a 54-year-old woman

20 months after replacement of the aortic root, ascending
aorta (A), aortic arch, innominate artery (I), and left common carotid artery for a type A dissection. Sagittal reconstruction from multidetector CT demonstrates disruption
of the distal anastomosis with contrast leaking (L) into a
contained mediastinal hematoma.
Fig. 20. Collapsed native aorta (arrow) medial to a descending aortic graft.
Cardiac valve reconstruction and replacement

Surgical methods for improving the function of
diseased cardiac valves include valve reconstruction
(valvuloplasty) and replacement with either mechanical prosthesis (ball-in-cage, single tilting disc, and
bileaflet prostheses), biologic prosthesis, or homograft (donor) valves. Radiologists should be familiar
with the radiographic appearance of the various valve
prostheses and the role of imaging in the detection of
complications [60].
Fig. 21. Bovine pericardial wrap in a 53-year-old man

after ascending aorta, arch, and descending thoracic aortic
graft placement. The graft is normal. Low-attenuation
material surrounding the left lateral aspect of the descending
aortic graft represents a bovine pericardial wrap (arrow).
557
Fig. 22. (A, B) Multiplanar CT reformatted images of an aortic stent graft in a 69-year-old woman placed for a type B dissection.
Displayed on soft tissue and bone window settings. Note the thrombosed false lumen (arrows).
Annuloplasty rings (Fig. 23) may be used to

restore the size and shape of the valve orifice during
valvuloplasty, examples of which include the Carpentier-Edwards ring (Baxter Health Care, Santa Ana,
California) and the Duran ring (Medtronic, Minneapolis, Minnesota). Both are radiopaque, with a gap
incorporated to provide more flexibility.
The radiographic appearances of the prosthetic
valves commonly encountered in clinical practice
are illustrated in Figs. 24 to 30. The St. Jude Medical
bileaflet valve (Fig. 26) is a widely used prosthetic
mitral and aortic valve. The leaflets are impregnated
with tungsten for radiopacity. The most common
radiographic appearance is a straight radiopaque line
representing one leaflet, with the other blurred

by motion. The Carpentier-Edwards bioprosthesis
(Fig. 30) and the Hancock porcine prosthesis, both
made of porcine aortic valves, are two of the biologic
prostheses currently in large-scale use. The struts of
the Carpentier-Edwards bioprosthesis are radiopaque,
whereas the alloy base ring is the only radiopaque
component of the Hancock porcine prosthesis. Cardiac valve homografts are not visible radiographically unless calcification occurs. This calcification is
usually limited to the donor aortic valve, producing
an eggshell appearance. Complications of prosthetic
valves include stenosis, endocarditis, structural failure, and periprosthetic leaks. Structural failure is
Fig. 23. Mitral (M) and tricuspid (T) valvuloplasty rings demonstrated on posteroanterior (A) and lateral (B) radiographs.
558
Fig. 24. (A, B) Posteroanterior and lateral views of a ball-in-cage type of aortic mechanical prosthesis (arrow).
very uncommon with the currently used mechanical

valves (Figs. 31 and 32). Echocardiography and cine
fluoroscopy play complementary roles in the evaluation of the performance and integrity of prosthetic
valves [61].
Cardiac transplantation
Cardiac transplantation is used for end-stage cardiomyopathy and coronary artery disease. In orthotropic cardiac transplantation the recipients heart is
removed through a median sternotomy. A cuff of both
native atria and the severed ends of the ascending
aorta and main pulmonary artery are retained. The
donor heart is joined to the recipients atria, aorta, and
pulmonary artery. On chest radiographs the normal
postoperative appearance typically includes an enlarged cardiac silhouette usually caused by a discrepancy between the size of the transplanted heart and
the native pericardium [62]. The size of the cardiac
silhouette decreases over time. A double right atrial
contour caused by overlap of the donor and recipient
right atria may be seen on postoperative chest radiographs. At CT, the normal postoperative appearance
of the heart and great vessels may include a high
redundant main pulmonary artery, a space between
the recipient superior vena cava and donor ascending
aorta, and a caliber change from the recipient to
donor ascending aorta [63].
Heterotropic cardiac transplantation is reserved
for patients with high pulmonary resistance who
receive a small donor organ, or who have acute or
Fig. 25. Medtronic-Hall (Medtronic, Minneapolis, MN) tilting disc mechanical aortic valve (arrow) demonstrated on posteroanterior (A) and lateral (B) radiographs.
Fig. 26. St Judes mitral valve (St Jude Medical, St Paul,

MN) on a lateral radiograph demonstrates the valve (arrow) in the open position. (Courtesy of P.A. Cascade, Ann
Arbor, MI.)
potentially reversible myocardial dysfunction. In heterotropic transplantation the donor heart is placed in
the right thoracic cavity and connected to the recipients heart in such a manner that the native right
ventricle provides most of the right-sided cardiac
output and the donor left ventricle provides the bulk
of the left-sided cardiac output. An enlarged cardiac
559
silhouette is seen on postoperative radiographs, with

the donor heart in the right hemithorax lateral to the
patients native heart.
The major complications that limit survival in
cardiac transplant recipients are infection, acute rejection, accelerated atherosclerosis, and malignancy.
Chest radiography and CT play an important role in
the diagnosis of infection and malignancy. Infection
screening is the most common indication for postoperative imaging in the cardiac transplant population,
with the lungs being the most common site of
infection. Bacteria are the most common pathogens
in the early postoperative period, whereas opportunistic pathogens, such as cytomegalovirus and Aspergillus species, predominate 2 to 6 months after
transplantation [63].
Radiologic detection of single or multiple pulmonary nodules or masses in a cardiac transplant recipient usually denotes the presence of infection or
malignancy [64]. Posttransplant lymphoproliferative
disorder occurs in 2% to 6% of cardiac transplant
recipients and manifests as either a solitary pulmonary
nodule or mass, multiple nodules or masses, and hilar
lymph node enlargement [65]. A high prevalence of
bronchogenic carcinoma has been shown in cardiac
transplant recipients, estimated at 1% to 2%, with a
median time to diagnosis of 34 months after surgery
[66]. Currently, endomyocardial biopsy and coronary
angiography are the standard methods for diagnosing
acute rejection and accelerated atherosclerosis.
Fig. 27. (A, B) Mosaic mitral bioprosthesis (Medtronic, Minneapolis, MN). The mosaic bioprosthesis is a stented new-generation
porcine heart valve for implantation in the aortic and mitral positions. Note the ringlike markers at the prosthesis (arrowheads).
560
Fig. 28. (A, B) Mosaic aortic bioprosthesis (arrowheads).
Congenital heart disease

The survival rate of patients after corrective surgery for congenital heart disease has significantly
improved, leading to an increased number of patients
being followed over extended periods of time and
into adulthood. The normal postoperative radiographic changes following palliative and corrective
procedures performed for the common congenital
conditions are discussed. MR imaging is a useful

comprehensive examination for the postsurgical
evaluation of patients with congenital heart disease,
providing both morphologic and functional information [67], whereas ECG-gated multidetector CT provides excellent morphologic information.
Isolated ostium secundum atrial septal defects are
usually closed either by direct suturing or placing a
pericardial patch. Transcatheter occluding devices
Fig. 29. (A) Posteroanterior and (B) lateral radiographs of a Bioprosthetic porcine mitral valve (arrows) (Baxter Health Care,
Irvine, CA).
561
Fig. 30. (A) Posteroanterior and (B) lateral radiographs of a Carpentier-Edwards aortic valve bioprosthesis (arrows) (Baxter
Health Care, Irvine, CA).
may be used for moderately sized septal defects.

The increased peripheral shunt vascularity caused
by the left-to-right shunt recedes almost immediately following closure, and the size of the heart
decreases over time. The dilatation of the main pulmonary artery, however, usually shows little if any
regression [68].
Fig. 31. Strut fracture of a Bjork-Shiley mechanical valve.

Posteroanterior radiograph demonstrates acute pulmonary
edema and an embolized strut in the upper abdomen (arrow).
Systemic to pulmonary arterial shunts are performed for palliation of right-to-left shunts in cyanotic congenital heart disease. These include the
subclavian artery to pulmonary artery shunt (Blalock-Taussig); the ascending aorta to right pulmonary
artery shunt (Waterston-Cooley); descending aorta to
left pulmonary artery (Fig. 33); and the superior vena
cava to right pulmonary artery anastomosis. Creation
of these shunts is accompanied by an increase in
pulmonary vascularity (right side in the WaterstonCooley shunt and bilateral in the Blalock-Taussig
shunt), and enlargement of the cardiac silhouette on
chest radiographs [68]. Unilateral rib notching on the
same side of the anastomosis is a recognized feature
of the Blalock-Taussig operation. More recent procedures for anastomosis of the subclavian artery to the
pulmonary artery use a Gortex graft without interruption of the subclavian artery. In this situation rib
notching should not occur [69].
Tetralogy of Fallot is the most common form of
complex congenital heart disease. Complete correction often requires widening of the outflow tract of the
right ventricle with a patch graft. Aneurysmal dilatation of this patch is not an uncommon complication,
producing a bulge on the left heart border just below
the hilum [70]. It is important to assess residual
anatomic problems, such as a residual ventricular
septal defect, the extent of pulmonary stenosis,
amount of pulmonary regurgitation, and biventricular
function in the follow-up of these patients.
Coarctation of the aorta may be treated surgically
with resection and anastomosis, patch grafts or a tube
graft, and using catheter-based techniques [71]. With
562
Fig. 32. Mechanical mitral valve dysfunction. (A) Frontal radiograph demonstrates massive cardiac enlargement. The left atrium
is enlarged secondary to prosthetic mitral valve dysfunction. (B) CT demonstrates a massive left atrium (LA) containing
thrombus (T).
successful eradication of the gradient after treatment,

the left border of the mediastinum remains abnormal
with at least partial obscuration of the aortic knob
[69]. During follow-up it is important to recognize
complications, such as pseudoaneurysms and recoarctation. A combination of spin echo and phase-contrast
MR imaging is used for postsurgical evaluation to
evaluate the functional severity of any recoarctation
and to quantify collateral flow [67]. The information
obtained by MR imaging obviates the need for

invasive angiography.
Summary
The normal postoperative appearances following
cardiac surgery and the imaging of the common
complications have been described. Awareness of
local surgical preferences and postoperative protocols
along with a teamwork approach with referring clinicians is emphasized.
References
Fig. 33. Status post Potts procedure in a 48-year-old woman

with tricuspid atresia. Coronal three-dimensional gadolinium-enhanced MR image demonstrates narrowing of the
Potts anastomosis (arrow) between the aorta and the left
pulmonary artery.
[1] Society of Thoracic Surgeons. National database. Chicago (IL): Society of Thoracic Surgeons; 2002.
[2] Leong CS, Cascade PN, Kazerooni EA, Bolling SF,
Deeb GM. Bedside chest radiography as part of a
postcardiac surgery critical care pathway: a means of
decreasing utilization without adverse clinical impact.
Crit Care Med 2000;28:383 8.
[3] OBrien W, Karski JM, Cheng D, Carroll-Munro J,
Peniston C, Sandler A. Routine chest roentgenography on admission to intensive care unit after heart operations: is it of any value? J Thorac Cardiovasc Surg
1997;113:130 3.
[4] Silverstein DS, Livingston DH, Elcavage J, Kovar L,
Kelly KM. The utility of routine daily chest radiography
in the surgical intensive care unit. Journal of TraumaInjury Infection & Critical Care 1993;35:643 6.
[5] ACR. American College of Radiology Appropriateness Criteria: thoracic. Routine daily portable X-ray.
Reston, VA: American College of Radiology; 1999.

[6] Salenger R, Gammie J, VanderSalm T. Postoperative
care of cardiac surgical patients. In: Cohn L, Edmunds
L, editors. Cardiac surgery in the adult. 2nd edition.
New York: McGraw-Hill; 2003. p. 453 78.
[7] Reyes A, Vega G, Blancas R, et al. Early vs conventional extubation after cardiac surgery with cardiopulmonary bypass [comment]. Chest 1997;112:193 201.
[8] Scott WL. Complications associated with central venous catheters: a survey [comment]. Chest 1988;94:
1221 4.
[9] Shin B, Ayella RJ, McAslan TC. Pitfalls of Swan-Ganz
catheterization. Crit Care Med 1977;5:125 7.
[10] Landay MJ, Mootz AR, Estrera AS. Apparatus seen on
chest radiographs after cardiac surgery in adults. Radiology 1990;174:477 82.
[11] Leitman BS, Naidich DP, McGuinness G, McCauley
DI. The left atrial catheter: its uses and complications.
Radiology 1992;185:611 2.
[12] Waldo AL, Henthorn RW, Plumb VJ. Temporary epicardial wire electrodes in the diagnosis and treatment
of arrhythmias after open heart surgery. Am J Surg
1984;148:275 83.
[13] Hartnell GG, Spence L, Hughes LA, Cohen MC,
Saouaf R, Buff B. Safety of MR imaging in patients
who have retained metallic materials after cardiac surgery. AJR Am J Roentgenol 1997;168:1157 9.
[14] Salenger R, Gammie J, VanderSalm T. Postoperative
care of cardiac surgical patients. In: Cohn L, Edmunds
L, editors. Cardiac surgery in the adult. 2nd edition.
New York: McGraw-Hill; 2003. p. 439 69.
[15] Creswell LL, Rosenbloom M, Cox JL, et al. Intraaortic
balloon counterpulsation: patterns of usage and outcome in cardiac surgery patients. Ann Thorac Surg
1992;54:11 8 [discussion: 18 20].
[16] Cascade PN, Rubenfire M, Kantrowitz A. Radiographic
aspects of the phase-shift balloon pump. Radiology
1972;103:299 302.
[17] Patel JJ, Kopisyansky C, Boston B, Kuretu ML,
McBride R, Cohen M. Prospective evaluation of complications associated with percutaneous intraaortic balloon counterpulsation. Am J Cardiol 1995;76:r1205 7.
[18] Barnett MG, Swartz MT, Peterson GJ, et al. Vascular
complications from intraaortic balloons: risk analysis.
J Vasc Surg 1994;19:81 7 [discussion: 87 9].
[19] Hyson EA, Ravin CE, Kelley MJ, Curtis AM. Intraaortic counterpulsation balloon: radiographic considerations. AJR Am J Roentgenol 1977;128:915 8.
[20] Cascade PN, Meaney JF, Jamadar DA. Methods of
cardiopulmonary support: a review for radiologists.
Radiographics 1997;17:1141 55.
[21] Muehrcke DD, McCarthy PM, Stewart RW, et al. Extracorporeal membrane oxygenation for postcardiotomy cardiogenic shock. Ann Thorac Surg 1996;61:
684 91.
[22] Jamadar DA, Kazerooni EA, Cascade PN, Fazzalari
FL, Vydareny KH, Bartlett RH. Extracorporeal membrane oxygenation in adults: radiographic findings and
correlation of lung opacity with patient mortality. Radiology 1996;198:693 8.
563
[23] Gross GW, Cullen J, Kornhauser MS, Wolfson PJ.

Thoracic complications of extracorporeal membrane
oxygenation: findings on chest radiographs and sonograms. AJR Am J Roentgenol 1992;158:353 8.
[24] Goldstein DJ, Oz MC. Mechanical support for postcardiotomy cardiogenic shock. Semin Thorac Cardiovasc
Surg 2000;12:220 8.
[25] Salzberg S, Lachat M, Zund G, et al. Left ventricular
assist device as bridge to heart transplantation: lessons
learned with the MicroMed DeBakey axial blood flow
pump. Eur J Cardiothorac Surg 2003;24:113 8.
[26] Peterson LR, McKenzie CR, Ludbrook PA, et al. Value
of saphenous vein graft markers during subsequent
diagnostic cardiac catheterization. Ann Thorac Surg
1999;68:2263 6.
[27] Milne EN. A physiological approach to reading critical
care unit films. J Thorac Imaging 1986;1:60 90.
[28] Henry DA. Radiological evaluation of the patient
after cardiac surgery. Radiol Clin North Am 1996;34:
119 35.
[29] Messent M, Sullivan K, Keogh BF, Morgan CJ, Evans
TW. Adult respiratory distress syndrome following
cardiopulmonary bypass: incidence and prediction.
Anaesthesia 1992;47:267 8.
[30] Katzberg RW, Whitehouse GH, deWeese JA. The
early radiologic findings in the adult chest after cardiopulmonary bypass surgery. Cardiovasc Radiol 1978;1:
205 15.
[31] Zylac C, Littleton J, Durizch M. Illusionary consolidation of the left lower lobe: a pitfall of portable radiography. Radiology 1988;167:653 5.
[32] Gaynes R, Bizek B, Mowry-Hanley J, Kirsh M. Risk
factors for nosocomial pneumonia after coronary artery bypass graft operations [comment]. Ann Thorac
Surg 1991;51:215 8.
[33] Gillinov AM, Davis EA, Alberg AJ, Rykiel M, Gardner TJ, Cameron DE. Pulmonary embolism in the
cardiac surgical patient. Ann Thorac Surg 1992;53:
988 91.
[34] Nelson RM, Jenson CB, Smoot III WM. Pericardial
tamponade following open-heart surgery. J Thorac Cardiovasc Surg 1969;58:510 6.
[35] Ellis K, Malm JR, Bowman Jr FO, King DL. Roentgenographic findings after pericardial surgery. Radiol
Clin North Am 1971;9:327 41.
[36] Bateman T, Gray R, Chaux A, et al. Right atrial
tamponade complicating cardiac operation: clinical, hemodynamic, and scintigraphic correlates. J Thorac Cardiovasc Surg 1982;84:413 9.
[37] Fyke III FE, Tancredi RG, Shub C, Julsrud PR, Sheedy
II PF. Detection of intrapericardial hematoma after
open heart surgery: the roles of echocardiography
and computed tomography. J Am Coll Cardiol 1985;
5:1496 9.
[38] Kay HR, Goodman LR, Teplick SK, Mundth ED. Use
of computed tomography to assess mediastinal complications after median sternotomy. Ann Thorac Surg
1983;36:706 14.
[39] Glanz S, Ravin CE, Deren MM. Benign pneumoperi-
564
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]

toneum following median sternotomy incision. AJR
Greenwald LV, Baisden CE, Symbas PN. Rib fractures
in coronary bypass patients: radionuclide detection.
Radiology 1983;148:553 4.
Boiselle PM, Mansilla AV. A closer look at the midsternal stripe sign. AJR Am J Roentgenol 2002;178:
945 8.
Boiselle PM, Mansilla AV, Fisher MS, McLoud TC.
Wandering wires: frequency of sternal wire abnormalities in patients with sternal dehiscence. AJR Am J
Roentgenol 1999;173:777 80.
Templeton PA, Fishman EK. CT evaluation of poststernotomy complications. AJR Am J Roentgenol 1992;
159:45 50.
Jolles H, Henry DA, Roberson JP, Cole TJ, Spratt JA.
Mediastinitis following median sternotomy: CT findings. Radiology 1996;201:463 6.
Salit IE, Detsky AS, Simor AE, Weisel RD, Feiglin D.
Gallium-67 scanning in the diagnosis of postoperative
sternal osteomyelitis: concise communication. J Nucl
Med 1983;24:1001 4.
Engle M, Ito T. The post pericardiotomy syndrome.
Am J Cardiol 1961;7:73 82.
Ball JB, Morrison WL. Experience with cardiac tamponade following open heart surgery. Heart Vessels
1996;11:39 43.
Miyake H, Iwanaga S, Suzuki K, et al. Delayed appearance of localized pericardial effusion after cardiac
surgery. AJR Am J Roentgenol 1991;157:19 20.
Kutcher MA, King III SB, Alimurung BN, Craver JM,
Logue RB. Constrictive pericarditis as a complication
of cardiac surgery: recognition of an entity. Am J Cardiol 1982;50:742 8.
White R. Magnetic resonance imaging in myocardial
and pericardial disease. In: Higgins C, Ingwall J,
Pohost G, editors. Current and future applications of
magnetic resonance in cardiovascular disease. Armonk
(NY): Futura; 1998. p. 97 103.
Soulen R, Grosh J. Cardiovascular system. In: Schwartz
E, editor. The radiology of complications in medical
practice. Baltimore: University Park; 1984. p. 186 7.
Thorsen MK, Goodman LR, Sagel SS, Olinger GN,
Youker JE. Ascending aorta complications of cardiac
surgery: CT evaluation. J Comput Assist Tomogr 1986;
10:219 25.
Ropers D, Ulzheimer S, Wenkel E, et al. Investigation
of aortocoronary artery bypass grafts by multislice spiral computed tomography with electrocardiographicgated image reconstruction. Am J Cardiol 2001;88:
792 5.
Gilkeson RC, Markowitz AH, Ciancibello L. Multisection CT evaluation of the reoperative cardiac surgery patient. Radiographics 2003;23:S3 17.
Doyle M, Spizarny D, Baker D. Saphenous vein graft
aneurysms after coronary artery bypass surgery. AJR
[56] Hughes MM, Rice TW, Simpfendorfer C. Aneurysmal

saphenous vein graft presenting as an anterior mediastinal mass. Catheter Cardiovasc Diagn 1991;24:265 7.
[57] Quint LE, Francis IR, Williams DM, Monaghan HM,
Deeb GM. Synthetic interposition grafts of the thoracic
aorta: postoperative appearance on serial CT studies.
Radiology 1999;211:317 24.
[58] Posniak HV, Demos TC, Marsan RE. Computed tomography of the normal aorta and thoracic aneurysms.
[59] Rofsky NM, Weinreb JC, Grossi EA, et al. Aortic aneurysm and dissection: normal MR imaging and CT
findings after surgical repair with the continuoussuture graft-inclusion technique. Radiology 1993;186:
195 201.
[60] Landay MJ, Estrera AS, Bordlee RP. Cardiac valve
reconstruction and replacement: a brief review. Radiographics 1992;12:659 71.
[61] Steiner RM, Mintz G, Morse D, et al. The radiology
of cardiac valve prostheses. Radiographics 1988;8:
277 98.
[62] Henry DA, Corcoran HL, Lewis TD, Barnhart GR,
Szentpetery S, Lower RR. Orthotopic cardiac transplantation: evaluation with CT. Radiology 1989;170:
343 50.
[63] Knisely BL, Mastey LA, Collins J, Kuhlman JE. Imaging of cardiac transplantation complications. Radiographics 1999;19:321 39 [discussion: 340].
[64] Haramati LB, Schulman LL, Austin JH. Lung nodules
and masses after cardiac transplantation. Radiology
1993;188:491 7.
[65] Armitage JM, Kormos RL, Stuart RS, et al. Posttransplant lymphoproliferative disease in thoracic organ
transplant patients: ten years of cyclosporine-based immunosuppression. J Heart Lung Transplant 1991;10:
877 86 [discussion: 886].
[66] Johnson WM, Baldursson O, Gross TJ. Double jeopardy: lung cancer after cardiac transplantation. Chest
1998;113:1720 3.
[67] De Roos A, Roest AA. Evaluation of congenital heart
disease by magnetic resonance imaging. Eur Radiol
2000;10:2 6.
[68] Curry G, Victoria B, Daicoff G. Radiological changes
following repair and palliation of right-to-left and admixture shunts. Radiol Clin North Am 1971;9:177 91.
[69] Barron M. Plain film diagnosis of common cardiac
anomalies in the adult. Radiol Clin North Am 1999;
37:416 8.
[70] Ascuitto RJ, Ross-Ascuitto NT, Markowitz RI, et al.
Aneurysms of the right ventricular outflow tract after
tetralogy of Fallot repair: role of radiology. Radiology
1988;167:115 9.
[71] Laks H, Marelli D, Plunkett M, Odim J, Myers J. Adult
congenital heart disease. In: Cohn L, Edmunds L, editors. Cardiac surgery in the adult. 2nd edition. New
York: McGraw Hill; 2003. p. 1335 8.
Radiol Clin N Am 42 (2004) 565 585
CT and MR imaging of the thoracic aorta: current

techniques and clinical applications
Servet Tatli, MD*, E. Kent Yucel, MD, Martin J. Lipton, MD
Division of Cardiovascular Imaging, Department of Radiology, Brigham and Womens Hospital, Harvard Medical School,
75 Francis Street, Boston, MA 02115, USA
Disease of the thoracic aorta can present with a

broad clinical spectrum of symptoms and signs. Patients may be rushed to the emergency room with
life-threatening complications of aortic trauma or dissection. They may be referred to outpatient clinics
with minimal complaints or be totally asymptomatic.
Furthermore, congenital lesions, such as aortic coarctation or vascular ring, may present at any age and
require a definitive diagnosis.
The accepted diagnostic gold standard, selective
digital subtraction angiography, is now being challenged by state-of-the-art CT angiography (CTA) and
MR angiography (MRA). These cross-sectional imaging techniques require no direct intra-arterial injection
of contrast material to evaluate the aorta, thereby
eliminating the catheter-induced complications from
invasive angiography. Currently, in many centers
cross-sectional imaging modalities are being used as
the first line of diagnosis to evaluate the cardiovascular system, and conventional angiography is reserved
for therapeutic intervention. Understanding the principles of CTA and MRA techniques is essential to
acquire diagnostic images consistently. Although all
imaging techniques are continuing to evolve, this
article reviews current CTA and MRA methods used
in the evaluation of thoracic aortic disease.
Imaging techniques
The entire aorta from the arch to iliac bifurcation
can be imaged in seconds with current multidetector
E-mail address: statli@partners.org (S. Tatli).
CT technology. Contrast-enhanced MRA (CE-MRA)

of the thoracic aorta can also be performed in just
over 20 seconds. Faster scanners have decreased the
acquisition time in a range of one breath-hold resulting in less motion from breathing. The acquired
imaging data set has high spatial and longitudinal
resolution allowing optimal three-dimensional reconstruction in selected projections, which is easier for
the radiologist and the referring physician or surgeon
to comprehend and evaluate. In addition to being
much safer and faster, CT and MR imaging display
not only the lumen of the aorta, as in conventional
angiography, but they also can demonstrate the vessel
wall and surrounding mediastinal structures. In fact,
this very important advantage of cross-sectional
imaging has improved the detection, evaluation, and
understanding of many aortic diseases. For instance,
the diagnosis and treatment of intramural hematoma
only became possible with the clinical use of crosssectional imaging techniques.
CT angiography
Development of CT angiography
Spiral CT technique acquires data continuously as
the patient travels through the scanner gantry [1].
Since its introduction in the early 1990s, spiral CT
technology has improved substantially so that CTA
has become the modality of choice for most centers in
the evaluation of acute and chronic thoracic aortic
disease. The introduction of four-detector scanners in
1998 with a faster gantry rotation time (0.5 seconds)
improved scanning efficiency by nearly eightfold in
comparison with a single-detector scanner. Since
doi:10.1016/j.rcl.2004.03.005
566
S. Tatli et al / Radiol Clin N Am 42 (2004) 565585
then, CT technology has been improved even further

with the introduction of eight-detector and finally
16-detector scanners. Worldwide many institutions
either have a 16-detector scanner or expect to install
one in the future. These state-of-the-art CT systems
can simultaneously acquire up to 16 submillimeter
(0.625 0.75 mm) sections with gantry rotation time
of approximately 0.40 seconds. Previous limitations
of CT for evaluating the vascular system with oldergeneration scanners have now been eliminated. Motion artifacts from breathing are no longer a problem,
because high-resolution imaging of the entire aorta
can be obtained in a single breath-hold. Imaging of all
phases of contrast enhancement has become possible
during the administration of a single contrast agent
bolus. Thinner section thickness allows, for the first
time in the history of CT, isotropic voxels. Obtaining
such data sets is essential for optimum CTA to obtain high-resolution three-dimensional reconstruction
and other postprocessing displays, such as maximum projection reformatting and maximum intensity
projection (MIP) in any selected plane. This revolutionary improvement has opened a new era in the
history of imaging. Despite ionizing radiation [2]
and the nephrotoxicity of contrast agents, the technique is widely available, fast, cost-effective, and
efficient. CT currently is the most frequently used
modality in the evaluation of the thoracic aorta and
has high diagnostic accuracy for detection of aortic
pathology [3].
Multidetector CT techniques
Multidetector scanning requires an understanding
of the basic principles for optimum results. Pitch and
collimation are two important parameters of image
acquisition. The slice thickness is dependent on the
detector collimation. The smaller the collimation the
thinner is the available slice thickness. The quality of
the three-dimensional reconstructions is directly related to the thickness of the obtained axial slices;
obtaining the thinnest available slices is mandatory
for optimum maximum projection reformatting and
three-dimensional reconstructions. CT arteriograms
should be acquired with the thinnest available collimation (0.625 0.75 mm). Selection of thick slice
collimations increases volume coverage of the longitudinal anatomic field. Other additional acquisitions
(precontrast or delay images) can be acquired with
thicker collimations (1.5 mm or greater), however,
because these images provide sufficient resolution,
and usually are not used in further postprocessing.
Pitch is defined as the ratio of the table speed per
gantry rotation to the collimation. Older scanners
require the use of high pitch values (eg, 2) for the

greatest anatomic coverage and spatial resolution for
a single breath-hold scan. Sixteen-detector scanners
allow coverage of the entire thoracic aorta in a single
breath-hold, however, using submillimeter collimation with lower pitch values (less than 1.5).
Contrast-enhancement methods
CT angiography requires the intravenous injection
of iodinated contrast agents to display the vessel
lumen. Severe allergy to iodine is the only absolute
contraindication. Premedication with steroids starting
12 hours before the study can be used for these
patients. Renal insufficiency can be exacerbated with
iodine and is a relative contraindication. In many
institutions, serum creatinine level of 1.5 mg/dL is the
threshold. Right antecubital veins are the preferred
injection sites because dense contrast in the left
brachycephalic vein may cause artifacts, which can
obscure the evaluation of the aortic arch and the
proximal segments of the great vessels. Consistently
optimum intraluminal enhancement is of importance
for a good CTA study and acquisition of images
during peak aortic opacification yields the best result. Optimum contrast enhancement is difficult to
achieve because it is affected by many variables including cardiac output, patients positioning, rate of
injection, location of the injection, scan direction,
scan delay, body weight, and pharmacokinetics properties of the contrast material [4]. Faster scanners
have resulted in shorter scan duration and contrast
bolus duration, allowing the use of less contrast
agent. This advantage can be used to inject the same
amount of contrast at a higher flow rate to achieve a
greater luminal enhancement.
Shorter acquisition time and the length of the
contrast bolus, however, led to use of contrast timing
as an important issue [5]. To overcome this problem a
timing bolus scan [6] can be performed. This method
was initially developed and has long been applied for
electron-beam scanners. The timing bolus technique
uses a small test bolus of contrast agent (10 15 mL)
with sequential dynamic scanning at the level of
proximal descending aorta to determine the time
delay between injection and arrival of contrast agent
bolus in the target artery. A test bolus should be
administered at a rate equivalent to that used for the
actual CTA acquisition, and should always be followed by a 15- to 20-mL saline chaser to prevent
pooling of the contrast bolus within the intravenous
line and peripheral veins, which requires a dualheaded injector [7]. Disadvantages of this method include increased patient radiation dose and additional
contrast agent usage. In addition, no clear enhancement peak may be encountered depending on the
patients body habitus.
In the authors institution and many others, empirical timing for imaging of the thoracic aorta with a
delay time of 25 seconds after the administration of
125 mL of nonionic contrast agent is used, resulting
in excellent arterial enhancement in most patients.
The delay time should be increased slightly (5
10 seconds) in older patients, and also in those with
decreased cardiac output or known aortic aneurysm.
A second scan at 60 seconds may be useful to detect
late-enhancing vascular structures.
Electrocardiographic gating
Although less frequent with multidetector CT
scanning, the motion caused by transmitted cardiac
pulsation to the major arteries creates problems.
These pulsation artifacts are particularly pronounced
in the proximal ascending aorta and may frequently
mimic an intimal flap resulting in a false-positive
diagnosis of aortic dissection. This problem can be
avoided by ECG gating, which is available in new CT
scanners and is being used routinely for imaging the
thoracic aorta. ECG gating can be applied prospectively or retrospectively [8].
In prospective ECG triggering, the acquisition of
the axial images of the aorta is performed during a
selected period of the R-R interval, for example
60%. The operator can select this delay time manually. In retrospective gating, axial images are obtained
with simultaneous recording of the patients ECG
signal. After completion of the scanning, only the
data acquired during a predefined phase of the cardiac
cycle (generally the diastolic phase) are used for
image reconstruction.
Study protocol
Oral contrast agent is not given before CTA and
all image acquisitions are obtained with breath-holding. An initial nonenhanced scan of the whole thoracic aorta is obtained (collimation: 1.5 mm; slice
thickness: 5 mm; reconstruction interval: 5 mm). This
nonenhanced scan is important for proper planning of
the contrast-enhanced scan and also useful in the
evaluation of certain entities, such as intramural
hematoma, and endoleak after endoluminal stent
placement. The thinnest available collimation is not
necessary for this nonenhanced acquisition and 5 mm
reconstructions are usually sufficient. ECG-gated
contrast-enhanced scanning is then performed from
567
just above the arch to the diaphragm using the thinnest

available collimation (collimation: 0.75 mm; slice
thickness: 0.75 mm; reconstruction interval: 0.4 mm).
Because many thoracic aortic diseases involve the
abdominal aorta, the authors routinely obtain a delay
scan from the lung apices to the pubic symphysis at
60-second delay with acquisition parameters similar
to the nonenhanced scan.
Postprocessing methods
Postprocessing of the raw image data is as important as the actual image acquisition technique to
obtain optimum final diagnostic images. Once CTA
acquisition is complete, the data are reconstructed at
the thinnest available section thickness for that acquisition, so that small structures can be visualized
and optimum three-dimensional reconstructions can
be obtained from these axial source images. The section thickness is determined by the width of a single
detector channel with multidetector scanners, and
cannot be smaller than the single detector channel.
It is well established that a 50% overlapping reconstruction interval provides greater diagnostic confidence and accuracy, and better three-dimensional
reconstructions with reduced stair-step artifact
[9,10]. The reconstruction interval should be equal
to the half value of the actual section thickness for
optimum results. With current technology using the
thinnest available collimation (0.625 0.75 mm) and
reconstruction interval (0.40 mm), generally several
hundred axial images are produced for a routine
thoracic aorta imaging study. This large data set is
impractical for a reader to review. The value of
comprehensively viewing the axial images directly
besides the three-dimensional reconstructions is well
established to detect vascular pathology and nonvascular abnormalities [11,12]. Thicker axial sections are
reconstructed from the axial source images to improve the efficacy of interpretation. In the authors
institution, 3-mm axial sections are reconstructed
with 2-mm intervals.
In addition to standard axial sections, sophisticated
reconstruction techniques are useful for displaying
and comprehending complex vascular anatomy and
its relationship to adjacent organs. It is also useful
for evaluating the extent of disease, precise delineation of the origin of the arch arteries, and accurate
aortic diameter measurements. Most modern scanners have postprocessing software or postprocessing
can be performed on a commercially available workstation. New user-friendly workstations allow single
or double oblique MIP and maximum projection
reformatting regardless of the original plane sec-
568
tion. Because of near isometric voxel volume now

available, these reconstructions approach the resolution and quality of the original sections. In the authors
institution, oblique sagittal and coronal MIP reformations are routinely obtained with 3-mm section
thickness and 1.5-mm reconstruction interval. MIP
displays bone and contrast-filled structures preferentially, and other lower-attenuation structures are not
well visualized.
Volume rendering is perhaps currently the most
sophisticated and robust three-dimensional imageprocessing tool, which incorporates all the information from the acquired CT data into the resulting
three-dimensional image. In this technique, each
voxel within the data set is mapped for the degree of
density and a color is assigned to its attenuation
values. The advantage of volume rendering is that it
potentially retains all the information in the image and
requires no assumptions to be made concerning structure or surfaces within the data. Although, it is not
used routinely for evaluating the thoracic aorta for
every patient, volume rendering is a useful tool to
demonstrate anatomy of the aorta despite difficulties
at the aortic root.
MR angiography
MR angiography techniques
MR angiography of the thoracic aorta usually
requires a combination of several available MR
imaging methods, each of which has certain advantages and contributes to the diagnostic versatility of
the technique. CE-MRA is the most widely used
MRA method because it is rapid and robust. CEMRA provides projection images of the aorta similar
to conventional invasive angiography. Black-blood
MR imaging permits assessment of the vessel wall by
saturating the signal from the lumen. Phase-contrast
imaging provides functional information about the
flow. Gradient-echo cine images can demonstrate
aortic regurgitation in the presence of disease of the
ascending aorta. Time-of-flight MRA offers limited
clinical value and today is not being used in routine
clinical imaging of the thoracic aorta. The field continues to develop and many new exciting MRA
methods, such as temporally resolved CE-MRA,
and parallel imaging techniques (eg, sensitivity encoding and simultaneous acquisition of spatial harmonics) promise further improvement in acquisition
time and resolution. The following paragraphs describe the most commonly used MRA techniques.
Black-blood vascular imaging

In conventional spin echo MR imaging, blood
usually is low in signal intensity because of movement of spins between a pair of (90 and 180 degree)
slice-selective radiofrequency pulses. If blood flows
out of the plane of the section in the time interval
between successive radiofrequency pulses, the result
is absence of signal, called a signal flow void [13].
For better depiction of intraluminal or mural abnormality dedicated black-blood technique is preferred
because it provides better suppression of the signal
from flowing blood (Fig. 1A) [14]. It is substantially
less efficient in terms of scan time than fast spin echo
technique, however, because it is a sequential-slice
imaging sequence. The technique uses two magnetization-preparation inversion pulses to suppress the
signal in the vascular lumen. The black-blood effect
is produced by application of a non slice-selective
inversion pulse (suppress the signal in the imaging
volume) followed by a slice-selective inversion pulse
(restores the signal in the imaging slice). The same
delay (T1) between the second inversion pulse and
the imaging sequence is necessary to null the signal
from protons within blood. During this T1, the nulled
blood protons from outside the imaging slice replace
the protons of the blood within the imaging slice that
received the second restorative inversion pulse. This
T1 is followed by a fast spin echo sequence with an
echo train of 8 to 16. Using cardiac gating (TR equal
to one R-R interval and TE of 20 30 millisecond,
giving T1-weighted imaging) and breath-holding
optimizes the image quality. The black-blood effect
is maximized by imaging in a plane perpendicular to
the vessel of the interest.
Phase-contrast imaging
Phase-contrast imaging is a unique MR imaging
technique that measures blood flow and can be used
in many clinical applications to evaluate physiologic
properties of blood flow. In phase-contrast imaging,
the phase shift difference between the moving spins
in the blood and that of the surrounding stationary
tissue is compared by using a bipolar gradient,
allowing detection of blood flow velocity. Two scans
are acquired (flow-sensitive scan and a flow-compensated reference scan), which are automatically
subtracted from each other. The resulting data are
processed into two sets of images: magnitude (anatomic) and phase-contrast (velocity) (Fig. 2). In
phase-contrast images, the gray value of each pixel
represents velocity information of that pixel. Higher
569
Fig. 1. A 55-year-old man with type B aortic dissection. T1-weighted axial image (A) with black-blood technique shows
excellent suppression of the luminal blood signal and demonstration of intimal flap (arrow) in the descending aorta. Contrastenhanced MRA of the aorta with sagittal oblique source (B), subtracted (C), MIP (D), and axial reformation (E) images show a
dissecting intimal flap (arrows). The subtracted image (C) was obtained by subtracting the unenhanced mask image from the
contrast-enhanced source image (B) and demonstrates better suppression of the background signal. MIP image (D) allows overall
evaluation of the dissecting intimal flap (arrows), which extends from the aortic arch to the abdominal aorta.
570
Fig. 2. Phase-contrast imaging of the aorta. Magnitude (A) and phase (B) axial images display the ascending (AA) and
descending (DA) aorta at the level of pulmonary artery (arrow). Flow encoding was set from the superior to the inferior direction
and images were obtained with ECG gating. On the phase image (B), the ascending aorta appears black and the descending aorta
white because of the opposite direction of the flow in these arteries. The volume and velocity of the flow can be calculated with
available software.
flow velocities are represented by higher signal

intensities, whereas blood flowing in the opposite
direction to the flow-encoded gradient appears dark.
By knowing the cross-sectional areas of a vessel
(measured from anatomic images), blood flow volume
and velocity can be calculated quantitatively. Before
the acquisition, the operator actively chooses a velocity-encoding factor to display the necessary vasculature. The better the encoding velocity matches the real
velocity of the region of interest, the more precise the
measurement becomes [15]. Larger velocity encoding
values increase noise, whereas smaller velocity encoding values cause aliasing artifact. Flow measurements
are most precise if the imaging plane is perpendicular
to the vessel of interest and flow encoding is set to
through plane flow [15].
Dynamic contrast-enhanced MR angiography
This technique was first described in 1993 [16]
and is now the principal MR technique for evaluating
the thoracic aorta. CE-MRA uses the T1 shortening
effects of gadolinium-based contrast agent, so that
the blood appears bright regardless of flow patterns or
velocity. Signal enhancement and overall image quality of CE-MRA depends on the intra-arterial contrast
agent concentration. The synchronization of image
acquisition and arrival of the bolus of contrast agent
in the region of interest is crucial to obtain high image
quality. Signals are collected in an area called
k-space before the Fourier transformation. The
central lines of the k-space determine image contrast,
whereas the peripheral lines of k-space contain data

encoding primarily for spatial resolution. The collection of the central lines of k-space during the plateau
phase of arterial enhancement is essential for optimal CE-MRA [17]. Filling the central portion of
k-space during peak arterial transit results in selective
arterial enhancement.
In a standard MR imaging technique, the lines of
the k-space are filled sequentially by phase-encoding
gradient from either bottom to top or vice versa with
filling of the central lines in the middle of the acquisition time. In a centric order, filling starts from
the lines in the center and progresses to periphery.
Image data in the center of the k-space, which determines image contrast, are collected earlier in the
beginning of the acquisition. If the central portion of
the k-space is filled before or during the upslope of the
contrast arrival, severe ringing artifacts limit the
diagnostic use of the image. Images acquired too long
after peak arterial contrast, however, are frequently
obscured by the enhancement of veins and soft tissues
limiting the use of reformats.
Contrast-enhanced MRA is performed using a
three-dimensional T1-weighted gradient echo imaging sequence. T1 weighting is obtained by using
gradient spoiling, minimum TR and TE, and relatively high flip angle (35 40 degree) to maximize
the signal-to-noise ratio. A narrow bandwidth
(31.125 kHz) can be used to achieve images with
less noise and an increased signal-to-noise ratio.
Narrowing the bandwidth results in an increase in
the available minimum TE. Imaging time is increased;
however, it can be compensated by decreasing the

number of excitations by filling the rest of the k-space
with partial field-of-view or partial Fourier acquisition
techniques. A 256 256 matrix maximizes spatial
resolution with a cost of longer imaging time. Lowering phase encoding causes lower special resolution,
which may be problematic in the evaluation of small
branches. Use of a rectangular field-of-view is one
way of retaining spatial resolution, while lowering
acquisition time. Although it is dependent to some
degree on the patients body habitus a field-of-view of
30 20 cm is generally sufficient. Additionally, to
minimize imaging time, the imaging plan should be
optimized to the anatomy being studied (eg, sagittal
oblique for thoracic aorta, coronal for evaluating the
aortic arch).
The use of heavily T1-weighted sequences introduces the problem of background signal from fat;
hence, a precontrast mask image is acquired and
used to subtract the fat signal. Obtaining a mask
image before CE-MRA sequence is preferred for
optimum background suppression. Both mask and
contrast-enhanced images are obtained with breathholding and the mask image is subtracted from the
contrast-enhanced sequence before MIP reconstructions (Fig. 1B, C). Routine use of a second scan also is
helpful in recovering diagnostic information in the
event of an early scan or to detect late-enhancing
vascular structures. Most commonly, postprocessing
involves the use of a MIP algorithm to create a
projection image (Fig. 1D). Areas with poor flow
contrast, including the edges of the blood vessel and
small vessel with slow flow, may be obscured by
overlap with brighter stationary tissue [18]. The quality of the MIP can be improved substantially by
reducing the pixel size and suppression of the signal
of stationary tissues [19]. In addition to routine MIP
reconstruction, axial reformatting is also very helpful
for imaging the aorta and in the authors department is
routinely performed (Fig. 1E). The acquisition parameters used for CE-MRA of the aorta are summarized
in Table 1.
Breath-holding is especially important for thoracic
aortic imaging and significantly improves image
quality; however, this may not be possible in every
571
case. Supplemental oxygen and hyperventilation can

help to improve breath-holding capacity for up to
25 seconds in most patients [20]. In ill or uncooperative patients, in whom breath-holding is limited to
just a few seconds, normal shallow breathing often
works well and results in adequate diagnostic images.
Contrast volume, delivery rate, and delay time are
all important parameters to optimize image quality.
The authors use a total of 40 to 50 mL of contrast
agent at 2 to 2.5 mL/second infusion rates. The injection of contrast agent should be followed by the
injection of 20 mL of saline at the same flow rate as
the contrast.
Several methods are used for determining the
correct contrast injection timing. The easiest timing
method uses a fixed scan delay with a delay time of
24 seconds for imaging the thoracic aorta [21].
Timing errors can occur, however, especially in
patients with aortic aneurysms or a low cardiac output. More precise timing methods should be used.
A timing-bolus scan, automatic detection of contrast
bolus passage, or MR imaging fluoroscopy [20,22,23]
can be used to achieve correct timing. In the authors
experience, automatic detection has been unreliable.
Bolus timing or MR imaging fluoroscopy techniques
are equally useful and reliable techniques. One of
these techniques with which the operator feels comfortable should be chosen and then used routinely.
Clinical applications
Atherosclerotic disease of the aorta
Atherosclerosis is the commonest disease afflicting the arterial system. It may present as an acute or
chronic syndrome [24]. The atherosclerotic process
begins in childhood as fatty streaks in the intima of
arteries and usually develops for many decades
before cardiovascular complications occur [25]. The
progression of atherosclerosis is accelerated in the
presence of risk factors, such as aging, hypertension,
hypercholesterolemia, and smoking [26]. Atherosclerosis causes intimal thickening with the accumulation
of lipid-laden foam cells and proliferation of smooth
Table 1
Acquisition parameters of contrast-enhanced MR angiography of the thoracic aorta
FOV
Matrix
Bandwidth
Flip angle
NEX
K-space
Contrast
amount/rate
Saline
amount/rate
Delay
Acquisition
time
30 cm
256 256
31.125 kHz
35 40
0.5
Centric
40 mL/2.5 mL
20 mL/2.5 mL
25 s
21 s
Abbreviations: FOV, field-of-view; NEX, number of excitations.
572
muscle resulting in atheromatous plaques (atheroma).

Minimal intimal thickening can progress to sessile or
protruding atheroma. Ulceration, calcification, and
superimposed thrombi are other features of atheroma.
Although extensive atherosclerosis may rarely cause
clinically significant stenotic or occlusive disease in
the thoracic aorta, more importantly it can predispose
to the development of clinical entities, such as aneurysm, dissection, and penetrating atherosclerotic ulcer
with severe morbidity and mortality.
It has been shown that increasing plaque thickness
(4 mm or more), plaque ulcerations (2 mm or more),
and mobile components (eg, thrombi) are associated
with a higher risk of stroke [27,28]. Aortic atherosclerosis is also a marker of coronary artery disease
and can be used to estimate the presence of coronary
artery disease [29]. High sensitivity (90%) and positive predictive value (95%) have been found for presence of significant coronary artery stenosis in patients
with atheroma in the aortic wall [30].
Atherosclerotic lesions of the thoracic aorta have
recently been recognized as an important cause of
stroke and peripheral embolization [31]. Until the
1990s, carotid disease and atrial fibrillation were the
two major entities that dominated the clinical approach to patients with stroke and peripheral embolization [31]. Despite these two sources of emboli, in
40% of patients with stroke, no etiology could be
found (cryptogenic stroke) [32]. Many case-control
studies documented the association between aortic
atheroma and embolic phenomena [27,28,33]. The reported prevalence of aortic atheroma seen in stroke
patients (21% 27%) was about the same magnitude
as the prevalence of carotid disease (10% 13%) and
atrial fibrillation (18% 30%) [27,28]. Although calcification represents one manifestation of the atherosclerosis, high-risk plaques are often uncalcified
(lipid-laden, vulnerable plaque).
CT and MR imaging are valuable noninvasive
techniques to look for the source of emboli in the
thoracic aorta and proximal arch arteries [34,35].
Contrast-enhanced CT can readily detect atherosclerotic plaques (Fig. 3) with comparable sensitivity and
specificity compared with transesophageal echocardiography, which has been used for plaque imaging
[36]. CT and MR imaging provide complete imaging
of the thoracic aorta including blind spots for transesophageal echocardiography [36,37]. Promising
experiments have been conducted in terms of MR
imaging characterization of atherosclerotic plaques
using ultrasmall superparamagnetic iron oxide particles [38]. These blood pool agents accumulate in
the plaque with high macrophage tissue content.
The inflammatory activity and age of the plaque,
Fig. 3. Oblique sagittal MIP reformation of CTA of a patient

with a history of stroke shows atheromatous plaque with
ulceration (arrow) in the aortic arch.
which may be clinically relevant information in

terms of response to lytic therapy, can be displayed
with this technique.
Aortic aneurysm
An aneurysm is defined as dilatation of the outer
aortic diameter at least 50% greater than the expected
diameter. The lesser degree of dilatation generally is
referred to as ectasia. Most thoracic aortic aneurysms are secondary to atherosclerosis. Atherosclerotic aneurysms are typically fusiform in shape and
more common in the descending aorta with a high
incidence of concomitant abdominal aortic aneurysm
(Fig. 4). Most of these aneurysms have a significant
amount of mural thrombus and calcification. Focal
eccentric aneurysms caused by atherosclerosis may
occasionally be encountered especially in the descending aorta (Fig. 5). Aortic valvular disease can
cause aneurysmal dilation of the ascending aorta
or vice versa an aneurysm of ascending aorta may
result in aortic insufficiency because of dilation of the
aortic root. In Marfan syndrome, aneurysms most
commonly occur in the proximal portion of the
ascending aorta involving the aortic root resulting
in pear-shaped aorta (Fig. 6). Important imaging
features of aortic aneurysms are the maximum diameter, the length, and involvement of major branch
vessels. The aneurysmal thoracic aorta grows at an
average rate of 1 mm per year with a high risk for
natural complications (rupture or dissection) at 6 cm
for the ascending aorta and 7 cm for the descending
aorta [39]. CT or MR imaging surveillance is recommended and surgical repair is considered when thoracic aorta reach a diameter of 5 to 6 cm. Peripheral
embolization, branch vessel involvement with steno-
573
Fig. 4. A 73-year-old woman with atherosclerotic aneurysm. Axial T1-weighted image with black-blood technique (A) and
sagittal oblique reformation of contrast-enhanced MRA (B) show the aneurysm involves the distal ascending (AA) and proximal
descending (DA) thoracic aorta.
sis, or occlusion are other clinically significant complications of aneurysms. Multiplanar reformations are
sometimes useful to measure the true diameter of the
tortuous aorta. CE-MRA is a luminogram and should
not be used for diameter measurement. The current
examination and also all available previous studies
should always be reviewed and compared. The comparison should include the earliest studies not to miss
a slowly expanding aneurysm; serial measurements
should be made at the same location of the aneurysm.
In patients with ascending aortic aneurysm, MR cine
imaging of the left ventricular outflow tract or phasecontrast imaging of the aortic valve can be added to
Fig. 5. Axial CTA image of a 61-year-old man with a saccular

aneurysm (arrows). Note the aneurysm arises from the aortic
arch and contains thick circumferential mural thrombus.
the routine imaging protocol for evaluating any associated aortic valve regurgitation (Fig. 7).
Aortic dissection
Aortic dissection occurs when blood dissects into
the media of the aortic wall through an intimal tear,
producing an intimal flap by separation of the false
lumen from the true lumen [40]. It generally is
secondary to chronic hypertension [41]. In young
patients with aortic dissection, an underlying process,
such as connective tissue disorders (Marfan or EhlersDanlos syndromes), should be investigated [42,43].
The proximal ascending aorta and the descending
aorta just distal to the left subclavian artery are two
common sites for initiation of the dissection. Sudden
onset of severe chest or back pain is a characteristic
presenting symptom [44,45]. Aortic dissection involving the ascending (Stanford type A) is a surgical
emergency with high mortality and may be complicated by contained rupture into the pericardium
causing pericardial tamponade, involvement of coronary arteries causing acute myocardial ischemia, and
extension to the arch arteries compromising brain
perfusion (Fig. 8) [41,44]. Additionally, aortic valve
disruption may lead to aortic regurgitation leading to
congestive heart failure. Dissections arising distal to
the left subclavian artery (Stanford type B) are
usually treated medically with surgical intervention
restricted for patients with signs of aortic expansion
and persistent clinical symptoms (see Fig. 1). Paraplegia caused by spinal cord ischemia is a frequent
complication and observed in up to 30% after surgery
of type B dissections [44]. The detection of side
574
Fig. 6. A 32-year-old man with Marfan syndrome. Axial (A) and sagittal oblique reconstruction (B) images of CTA show a
marked dilation of the aortic root (arrows) at the level of the sinus of Valsalva resulting in a pear-shaped aortic root, which is
typical for this syndrome.
Fig. 7. A 35-year-old man with Marfan syndrome. Sagittal oblique image (A) from diastolic phase of steady-state free precession
cine acquisition nicely displays dilated aortic root (black arrows) with regurgitant flow caused by aortic valve insufficiency
(white arrows). Phase-contrast images from the level of aortic valve obtained during systole (B) and diastole (C). Systolic flow
from the left ventricle to aorta is black (arrow in B); however, the regurgitant flow from the aorta to ventricle (arrow in C) is
white because flow encoding was set in a superior-to-inferior direction.
575
Fig. 8. A 60-year-old man with a history of hypertension and family history of aortic dissection presented with a sharp epigastric
pain radiating to chest. CTA showed type A dissection. Axial image at the level of sinus of Valsalva (A) shows intimal flap (black
arrows) and tear site (black arrowhead). Note a large hemopericardium (white arrows). The intimal flap was involving the ostia
of right coronary artery (not shown here). Dissection of the descending aorta is also seen (white arrowhead). Axial image at the
level of the right pulmonary artery (B) clearly shows intimal flap separating the false lumen (arrows) from the true lumen
(arrowheads). Note low density of the false lumen because of delayed flow. (C) Axial image at more cranial level reveals
involvement of major arch arteries (arrows). Coronal image from three-dimensional reformation (D) nicely displays the extent of
the dissection (arrows).
branch involvement is not a first-line prerequisite for

surgery. The recently introduced interventional techniques, such as aortic fenestration and stent graft
implantation, have opened new therapeutic options,
which can be performed before or after surgery [40].
Multidetector contrast-enhanced CT is a valuable
imaging tool in the emergency setting to evaluate
clinically suspected dissection with excellent accuracy
[46]. It is widely available and fast. The extent of the
dissection, true and false lumen size, false lumen
patency, and branch vessel involvement can be evaluated by CT, which should extend from the neck base
to the aortic bifurcation. Periaortic fluid, pericardial
effusion, and pleural effusion are important additional

signs to alert the radiologist in terms of catastrophic
complications with high mortality.
MR imaging is the most sensitive method for
diagnosing aortic dissection and has the same specificity as CT [47] but its use is subject to availability
and limited in the emergency setting because of difficulties in handling emergency cases. It is reserved
for patients with allergy to iodine and renal failure. It
is the preferred tool for imaging chronic dissections
and postsurgical follow-up.
The demonstration of an intimal flap separating
two lumens is the key diagnostic finding. The con-
576
of the degree of the communication between the true

and false lumen (Fig. 9). Because of intraluminal high
pressure and wall stress, the false lumen has a
tendency to progress (especially in the absence of
an adequate re-entry tear) and enlarge over the time to
form an aneurysm with risk of rupture, and must be
closely followed.
Intramural hematoma
Fig. 9. Axial T1-weighted black-blood image of the aorta

at the hiatus shows aortic dissection with compressed true
lumen posteriorly. Note the false lumen is situated anteriorly
and partially thrombosed (arrow).
vexity of the intimal flap is generally toward the false

lumen, which surrounds the true lumen. The false
lumen usually has slower flow, greater cross-sectional
area, and may contain thrombi [48]. The intimal tears
(named as entry or re-entry sites) most of the time
can be demonstrated easily by CT or MR imaging
(see Fig. 8). It is important to localize the tear sites
because surgery and stent graft implantations usually aim at occluding the tear to induce thrombus
formation in the false lumen. The extent of the
thrombus in the false lumen is an indirect indicator
Intramural hematoma is an atypical form of dissection without flow in the false lumen or a discrete
intraluminal flap and constitutes 10% to 20% of acute
aortic syndromes [44,49,50]. Once considered an
entity diagnosed only at necropsy, with the introduction of high-resolution cross-sectional imaging in
clinical use, the in vivo diagnosis is now feasible.
Arterial hypertension is the most frequent predisposing factor as in aortic dissection [50]. The pathogenesis of intramural hematoma still remains unclear.
Spontaneous rupture of the aortic vasa vasorum or
penetrating atheromatous ulcer was proposed as the
initiating event [51]. Intramural hematoma most frequently involves the ascending aorta or proximal
segment of the descending aorta as in those with
classic dissection. The acute complications of aortic
dissection, such as aortic insufficiency, rupture into
pericardium, and branch vessel involvement, may
also occur with intramural hematoma. It may regress
over the time with resorption of the hematoma or
progress to develop serious complications [52,53].
It is generally considered to be a precursor of overt
Fig. 10. A 49-year-old man with a history of hypertension presented with acute chest pain. Precontrast (A) and postcontrast
(B) axial CTA images revealed intramural hematoma. Note a high-density crescent-shaped wall thickening in the descending
thoracic aorta (arrows), which is better appreciated on precontrast image.
dissection rather than separate clinical entity [54].

Although the clinical symptoms, prognostic impact of
the location, and its standard treatment have been
considered similar to those of classic aortic dissection
[50,52]. Recently, controversial reports have been
published in terms of prognosis and management of
intramural hematoma. There are reports of favorable
responses to medical treatment with complete resorption of the hematoma without surgical intervention
[53,55 57]. In a multicenter study, however, von
Kodolitsch et al [54] demonstrated that about 50%
of the cases with intramural hematoma complicated
by overt dissection, contained rupture or aneurysm
(more than 7 cm in diameter) within 30 days of initial
presentation. Kaji et al [56] found a maximum aortic
diameter greater than or equal to 5 cm to be an
independent predictor of progression and suggested
monitoring these patients with frequent imaging.
The diagnosis of intramural hematoma relies on
the visualization of intramural blood or evidence of
localized increased wall thickness. The high density
of fresh hematoma on unenhanced CT imaging is
specific for intramural hematoma (Fig. 10). In MR
imaging, it can be identified as crescentic thickening
of the aortic wall with high signal intensity in
T1-weighted images (Fig. 11). Fat saturation images
before contrast can be helpful in differentiating intramural hematoma from surrounding mediastinal fat.
Penetrating atherosclerotic ulcer
Penetrating atherosclerotic ulcer is a condition
characterized by ulceration of an atherosclerotic
plaque that penetrates through the intima into the
Fig. 11. Axial T1-weighted image with black-blood and

fat suppression shows crescent-shaped, high signal intensity wall thickening consistent with an intramural hematoma (arrow).
577
Fig. 12. A 61-year-old man with high blood pressure.

Sagittal oblique MIP image from contrast-enhanced MRA
shows a large outpouching (arrow) in the medial anterior
aspect of the descending aorta. Note also the diffuse aneurysmal dilation of the aorta.
media of the aortic wall [58]. A penetrating atherosclerotic ulcer is typically located in the descending
thoracic aorta and can be associated with a variable amount of hematoma within the aortic wall
[49,59 61]. It generally affects elderly individuals
with hypertension and extensive aortic atherosclerosis [60], presenting with chest or back pain. Penetrating atherosclerotic ulcer can result in localized
tear through the adventitia forming pseudoaneurysm,
which can be quite large (Fig. 12). Many penetrating
atherosclerotic ulcers are diagnosed in asymptomatic
patients who undergo imaging for other reasons and
remain unchanged over time. It can be complicated,
however, by saccular or fusiform aortic aneurysms,
classic dissection, or aortic rupture [60,62]. Intramural
hematoma may also result in a focal outpouching
resembling penetrating atherosclerotic ulcer (Fig. 13)
[49]. There is discrepancy in the prognosis of penetrating atherosclerotic ulcer on outcome studies. Tittle
et al [63] reports that rupture occurred during the
initial admission in 38% of cases, whereas others
indicated more benign course in most of the patients
[60,62,63].
In imaging, penetrating atherosclerotic ulcer is
seen as an outpouching extending beyond the contour
of the aortic lumen and CT and MR imaging can
demonstrate associated intramural hematoma in acute
578
Fig. 13. Unenhanced axial CT image (A) shows intramural hematoma (arrows) in the descending aorta. On follow-up CTA, a
focal outpouching (arrows) developed at this region, resembling penetrating atherosclerotic ulcer as seen on axial (B) and sagittal
oblique MIP (C) images.
stage. The diagnosis can be difficult when the presentation overlaps with atypical focal aortic dissection. In
fact, several other different entities (eg, focal aneurysm with irregular atherosclerotic thrombus, or
contained aortic rupture) may also produce ulcerlike
lesions in the aorta resembling penetrating atherosclerotic ulcer [62]. As in classic dissection, these lesions
are managed surgically if located in the ascending
aorta, whereas more distal penetrating atherosclerotic
ulcer without clinical signs of instability is managed
medically and followed by sequential imaging [64].
Unstable descending aorta-penetrating atherosclerotic
ulcer is considered for more aggressive treatment,
such as stent-graft placement. Endovascular stent graft
is becoming a popular method to treat this entity given

that the disease tends to occur in elderly patients at
high surgical risk because of other comorbidities [49].
Infectious and inflammatory aortic disease
Aortitis inflammation of the aorta can be secondary to infectious agents (spirochetes); connective
tissue disorders (systemic lupus); or unknown etiology (Takayasus and giant cell arteritis). Inflammation of the aorta can cause aortic dilation, resulting in
aortic insufficiency, fibrous thickening or ostial stenosis of major branches, resulting in reduced or
absent pulses, ocular disturbances, neurologic defi-
579
Fig. 14. A 41-year-old woman with Takayasus arteritis. Contrast-enhanced axial image with black-blood technique (A) from
the level of the aortic arch obtained with ECG gating, breath-holding, and fat saturation. Note enhancing thickening of the aortic
wall (arrows). Coronal MIP image from contrast-enhanced MRA (B) shows involvement of the aortic arch branches. Total
occlusion of left subclavian artery (white arrow), severe stenosis of the origin of the left common carotid (black arrow), and long
segment stenosis of the right common carotid artery (white arrowheads) are evident. Note also aberrant origin of left vertebral
artery from the arch (black arrowhead).
cits, and other manifestations of vascular impairment

depending on the artery involved.
Syphilitic aortitis predominantly involves the ascending aorta or aortic arch, leading to ascending
aortic aneurysms and aortic valve incompetence.
Asymmetric saccular involvement and heavy calcifications are typical. Coronary ostial stenosis may
cause angina.
Takayasus arteritis (pulselessness syndrome) occurs most frequently in young Asian women, although the disorder has been observed worldwide
[65]. It most often affects the thoracic aorta and
branches. Stenoses and occlusions are characteristic
Fig. 15. Coronal MIP image from contrast-enhanced MRA

of a patient with known giant cell arteritis. Note occlusion of
both subclavian arteries (arrows).
of the disorder, but dilatation and aneurysms are not

rare. Systemic signs and symptoms characterize the
early phase of the disease. The significant feature of
this phase is wall thickening of the aorta. CT and MR
imaging are highly sensitive and specific for the
detection of wall thickening, which enhances with
contrast agent (Fig. 14A). It usually takes years to
develop the occlusive complications of the late-phase
disease. MIP reformations of the MRA are very helpful in determining the degree and extent of occlusive
Fig. 16. A 39-year-old woman with a history of aortic

coarctation repair presented with blood pressure discrepancy
between the upper and lower extremities. Sagittal oblique
MIP reformation from CTA shows residual stenosis in the
proximal descending aorta (arrowheads). Note postoperative
pseudoaneurysm (arrow) adjacent to the coarctation.
580
disease of the advance disease in the aorta and arch

branches (Fig. 14B).
Giant cell arteritis (temporal arteritis) is another
arteritis of unknown origin affecting the primary and
secondary branches of the aorta and sometimes the
aorta itself [66]. It is the most common form of vasculitis in Whites, occurring most commonly in women and individuals over the age of 50 years. In
addition to occlusive disease of the branches of the

aorta, it sometimes may cause wall thickening and
aneurysmal dilation of the aorta (Fig. 15).
Adult congenital aortic disease
Although many patients with arch anomalies
present in childhood, others may not be recognized
Fig. 17. CTA of a patient with suspicion of aortic coarctation on plain chest radiographs. Sagittal oblique MIP (A,B) and threedimensional reconstruction with volume rendering technique (C) demonstrates that there is aortic tortuosity (arrow) rather than a
true coarctation (pseudocoarctation). The patient did not have any clinical or other evidence of a coarctation.
until adulthood. Aortic coarctation may present in the

adult population as a part of the work-up of hypertension or an enlarged thoracic aorta. More commonly,
however, CT or MR imaging is called on in the
postoperative evaluation of congenital cardiovascular
diseases, such as aortic coarctation (Fig. 16). Pseudocoarctation is redundancy and tortuosity of the
thoracic aorta without hemodynamic consequences
(Fig. 17). On occasion, it may be difficult to differentiate from true coarctation on routine imaging
alone. Phase-contrast flow imaging is helpful in this
condition. In aortic coarctation, the cross-sectional
flow through the aorta at a level just distal to the
coarctation and at the diaphragm can be measured
using phase-contrast MR imaging. Normally, the ratio
of distal to proximal flow is less than 1 because of
antegrade flow through the intercostal arteries. In
hemodynamically significant coarctation with retrograde flow from the intercostals arteries, the distal to
proximal flow ratio is greater than 1. Aberrant origin
of right subclavian artery is one of the common
malformations of the aortic arch and may be silent
until adult age unless aneurysmal or tortuous. The
right aortic arch with mirror-image branching is the
anatomic counterpart of normal left arch and has a
high incidence of associated congenital heart disease;
however, the right arch with aberrant left subclavian
artery (Fig. 18) is the most common right arch variant
found in adults [67].
Postsurgical evaluation of aortic diseases
Many symptomatic thoracic aortic diseases carry
high risk for mortality requiring surgical repair, which
Fig. 18. Axial CTA image through the upper chest showing a right-sided arch (arrows) with an aberrant left
subclavian artery arising from the diverticulum of Kommeral (arrowheads), which is coursing to left side posterior to
the trachea.
581
Fig. 19. Sagittal oblique MIP image from contrast-enhanced

MRA of a patient with a history of ascending aortic surgery
due to type A aortic dissection. Note interposed Dacron graft
(arrows) and replaced aortic valve (arrowheads).
commonly involves thoracotomy with graft interposition. Dacron (polyester) graft is the most commonly
used synthetic graft. The diseased native aorta can be
either completely replaced by the graft (interposition
technique) or wrapped around the inserted graft
(inclusion technique). In the postoperative period,
follow-up by CT or MR imaging is routinely recommended to identify graft stability and possible
complications, such as graft dehiscence or pseudoaneurysm. In addition, the coexisting disease of
the descending aorta, which commonly remains unrepaired unless symptomatic, requires surveillance in
terms of progress of the disease. The anastomosis site
can be identified by abrupt change in aortic caliber or
an abrupt transition between nonatherosclerotic (graft)
and atherosclerotic (native) aortic wall (Fig. 19).
High-attenuation rings (felt strips) can help to distinguish the proximal anastomosis site. These rings are
used to reinforce the proximal anastomosis. Lowattenuation or soft tissue density material surrounding
or adjacent to the aortic graft can be seen months or
years after the surgery. These materials possibly
represent an old hematoma that has evolved into
fibrous tissue and should not be mistaken for leakage
or infection. The use of an interposition graft requires
the coronary arteries to be anastomosed to the graft
with a button of native aortic root. These buttons can
582
Fig. 20. This patient was status post endovascular stent (arrows) replacement because of aneurysm. Axial (A) and sagittal
MIP reformation (B) of CTA show marked contrast leakage into the aneurysmal sac caused by malpositioning of the
stent (arrowheads).
occasionally be prominent and simulate a pseudoaneurysm at the proximal graft anastomosis.

The treatment of descending thoracic aortic aneurysms using an endovascular stent is a recent advancement and is receiving increasing attention as an
alternative to open surgical repair. These prostheses
have been increasingly used to treat aneurysms,
dissections, and traumatic ruptures of the descending
thoracic aorta. After the use of such stents, serial
imaging studies must be performed to detect device
failure, such as stent migration, failure, or endoleak
before the development of devastating clinical sequelae (Fig. 20) [68].
Traumatic aortic injuries

Because of its noninvasiveness and availability in
most emergency departments, CT has become the
main imaging modality in the evaluation of traumatic
aortic injuries. It allows rapid assessment of the entire
thorax and abdomen while patients are being monitored. In the evaluation of patients with acute thoracic
injuries, helical CT was shown to have excellent
sensitivity (100%) and negative predictive value
(100%) [69,70]. Intramural hematoma, intimal tear,
pseudoaneurysm, and extravasation of the contrast
material from the aorta are typical CT imaging find-
Fig. 21. CTA of a 22-year-old man who sustained injury from a motorcycle accident. Axial (A) and sagittal reformatted
(B) images elegantly display traumatic rupture of the descending aorta (black arrows). Note mediastinal hematoma (white arrow)
and bilateral pleural effusions (arrowheads).
ings of acute traumatic aortic injuries (Fig. 21). The

aortic isthmus is the most commonly injured aortic
site [71]. Aortic ductus diverticulum is an anatomic
variant and sometimes mimics a pseudoaneurysm.
Summary
CT angiography and MR angiography are valuable tools in the evaluation of acute and chronic disorders of the thoracic aorta. These noninvasive
modalities provide crucial information about the
vessel wall and surrounding mediastinal structures
in addition to aortic lumen and should be used as a
first-line diagnostic method, reserving conventional
angiography for therapeutic intervention. CT is fast,
robust, widely available, and cost-effective. MR
imaging uses no ionizing radiation or nephrotoxic
iodine and allows a comprehensive evaluation of the
aorta including functional flow measurements and
evaluation of aortic valve.
Understanding of principles of the techniques is
important to obtain consistently diagnostic images.
Proper demonstration of the obtained images by
means of postprocessing techniques is equally important to communicate properly with referring physicians. Knowing the natural history of aortic diseases
and imaging features can lead to an accurate diagnosis and proper management of patients with aortic disease.
References
[1] Kalender WA, Seissler W, Klotz E, Vock P. Spiral
volumetric CT with single-breath-hold technique,
continuous transport, and continuous scanner rotation.
Radiology 1990;176:181 3.
[2] Hidajat N, Maurer J, Schroder RJ, Wolf M, Vogl T,
Felix R. Radiation exposure in spiral computed tomography: dose distribution and dose reduction. Invest
Radiol 1999;34:51 7.
[3] Hartnell GG. Imaging of aortic aneurysms and dissection: CT and MRI. J Thorac Imaging 2001;16:35 46.
[4] Fleischmann D, Rubin GD, Bankier AA, Hittmair K.
Improved uniformity of aortic enhancement with customized contrast medium injection protocols at CT
angiography. Radiology 2000;214:363 71.
[5] Lipton MJ, Higgins CB. Computed tomography: the
technique and its use for the evaluation of cardiocirculatory anatomy and function. Cardiol Clin 1983;1:
457 71.
[6] van Hoe L, Marchal G, Baert AL, Gryspeerdt S, Mertens L. Determination of scan delay time in spiral
CT-angiography: utility of a test bolus injection.
J Comput Assist Tomogr 1995;19:216 20.
583
[7] Hopper KD, Mosher TJ, Kasales CJ, TenHave TR,

Tully DA, Weaver JS. Thoracic spiral CT: delivery
of contrast material pushed with injectable saline solution in a power injector. Radiology 1997;205:269 71.
[8] Kachelriess M, Kalender WA. Electrocardiogram-correlated image reconstruction from subsecond spiral
computed tomography scans of the heart. Med Phys
1998;25:2417 31.
[9] Urban BA, Fishman EK, Kuhlman JE, Kawashima A,
Hennessey JG, Siegelman SS. Detection of focal hepatic lesions with spiral CT: comparison of 4- and
8-mm interscan spacing. AJR Am J Roentgenol
1993;160:783 5.
[10] Zeman RK, Silverman PM, Berman PM, Weltman
DI, Davros WJ, Gomes MN. Abdominal aortic aneurysms: evaluation with variable-collimation helical CT
and overlapping reconstruction. Radiology 1994;193:
555 60.
[11] Katz DS, Jorgensen MJ, Rubin GD. Detection and
follow-up of important extra-arterial lesions with helical CT angiography. Clin Radiol 1999;54:294 300.
[12] Kalender WA, Prokop M. 3D CT angiography. Crit
Rev Diagn Imaging 2001;42:1 28.
[13] Lipton M, Herfkens RJ, Higgins CB. Darstellung kardiovascularer strukturen mit hilfender kernspin-tomographie. Van Swieten Tagun 1983;37:39 42.
[14] Edelman RR, Mattle HP, Wallner B, Bajakian R, Kleefield J, Kent C, et al. Extracranial carotid arteries:
evaluation with black blood MR angiography. Radiology 1990;177:45 50.
[15] Lotz J, Meier C, Leppert A, Galanski M. Cardiovascular flow measurement with phase-contrast MR imaging:
basic facts and implementation. Radiographics 2002;
22:651 71.
[16] Prince MR, Yucel EK, Kaufman JA, Harrison DC,
Geller SC. Dynamic gadolinium-enhanced threedimensional abdominal MR arteriography. J Magn
Reson Imaging 1993;3:877 81.
[17] Svensson J, Petersson JS, Stahlberg F, Larsson EM,
Leander P, Olsson LE. Image artifacts due to a timevarying contrast medium concentration in 3D contrast-enhanced MRA. J Magn Reson Imaging 1999;
10:919 28.
[18] Anderson CM, Saloner D, Tsuruda JS, Shapeero LG,
Lee RE. Artifacts in maximum-intensity-projection
display of MR angiograms. AJR Am J Roentgenol
1990;154:623 9.
[19] Yucel EK, Anderson CM, Edelman RR, Grist TM,
Baum RA, Manning WJ, et al. AHA scientific statement: magnetic resonance angiography: update on
applications for extracranial arteries. Circulation 1999;
100:2284 301.
[20] Earls JP, Rofsky NM, DeCorato DR, Krinsky GA,
Weinreb JC. Breath-hold single-dose gadoliniumenhanced three-dimensional MR aortography: usefulness of a timing examination and MR power injector.
Radiology 1996;201:705 10.
[21] Prince MR, Chabra SG, Watts R, Chen CZ, Winchester
PA, Khilnani NM, et al. Contrast material travel times
584
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]

in patients undergoing peripheral MR angiography.
Radiology 2002;224:55 61.
Foo TK, Saranathan M, Prince MR, Chenevert TL.
Automated detection of bolus arrival and initiation of
data acquisition in fast, three-dimensional, gadoliniumenhanced MR angiography. Radiology 1997;203:
275 80.
Wilman AH, Riederer SJ, King BF, Debbins JP, Rossman PJ, Ehman RL. Fluoroscopically triggered contrast-enhanced three-dimensional MR angiography
with elliptical centric view order: application to the
renal arteries. Radiology 1997;205:137 46.
Lipton M. Computed tomography of thoracic aortic
disease. American College of Radiology Annual Meeting: Syllabus for the categorical course on cardiovascular imaging 1995;107 13.
Lipton MJ, Bogaert J, Boxt LM, Reba RC. Imaging of
ischemic heart disease. Eur Radiol 2002;12:1061 80.
Tunick PA, Rosenzweig BP, Katz ES, Freedberg RS,
Perez JL, Kronzon I. High risk for vascular events in
patients with protruding aortic atheromas: a prospective study. J Am Coll Cardiol 1994;23:1085 90.
Amarenco P, Cohen A, Tzourio C, Bertrand B, Hommel M, Besson G, et al. Atherosclerotic disease of
the aortic arch and the risk of ischemic stroke. N Engl
J Med 1994;331:1474 9.
Jones EF, Kalman JM, Calafiore P, Tonkin AM,
Donnan GA. Proximal aortic atheroma: an independent risk factor for cerebral ischemia. Stroke 1995;
26:218 24.
Witteman JC, Kannel WB, Wolf PA, Grobbee DE,
Hofman A, DAgostino RB, et al. Aortic calcified
plaques and cardiovascular disease (the Framingham
Study). Am J Cardiol 1990;66:1060 4.
Fazio GP, Redberg RF, Winslow T, Schiller NB. Transesophageal echocardiographically detected atherosclerotic aortic plaque is a marker for coronary artery
disease. J Am Coll Cardiol 1993;21:144 50.
Tunick PA, Kronzon I. Atheromas of the thoracic
aorta: clinical and therapeutic update. J Am Coll
Cardiol 2000;35:545 54.
Sacco RL, Ellenberg JH, Mohr JP, Tatemichi TK, Hier
DB, Price TR, et al. Infarcts of undetermined cause: the
NINCDS Stroke Data Bank. Ann Neurol 1989;25:
382 90.
The French Study of Aortic Plaques in Stroke Group.
Atherosclerotic disease of the aortic arch as a risk factor for recurrent ischemic stroke. N Engl J Med 1996;
334:1216 21.
Kutz SM, Lee VS, Tunick PA, Krinsky GA, Kronzon I.
Atheromas of the thoracic aorta: a comparison of transesophageal echocardiography and breath-hold gadolinium-enhanced 3-dimensional magnetic resonance
angiography. J Am Soc Echocardiogr 1999;12:853 8.
Krinsky GA, Kaminer E, Lee VS, Rofsky NM, Weinreb
JC. The effects of apnea on timing examinations for
optimization of gadolinium-enhanced MRA of the thoracic aorta and arch vessels. J Comput Assist Tomogr
1998;22:677 81.
[36] Tenenbaum A, Garniek A, Shemesh J, Fisman EZ,

Stroh CI, Itzchak Y, et al. Dual-helical CT for detecting
aortic atheromas as a source of stroke: comparison
with transesophageal echocardiography. Radiology
1998;208:153 8.
[37] Krinsky GA, Freedberg R, Lee VS, Rockman C,
Tunick PA. Innominate artery atheroma: a lesion seen
with gadolinium-enhanced MR angiography and often
missed by transesophageal echocardiography. Clin
Imaging 2001;25:251 7.
[38] Schmitz SA, Winterhalter S, Schiffler S, Gust R, Wagner S, Kresse M, et al. USPIO-enhanced direct MR
imaging of thrombus: preclinical evaluation in rabbits.
Radiology 2001;221:237 43.
[39] Elefteriades JA. Natural history of thoracic aortic
aneurysms: indications for surgery, and surgical versus nonsurgical risks. Ann Thorac Surg 2002;74:
S1877 80 [discussion: S1892 8].
[40] Erbel R. Diseases of the thoracic aorta. Heart 2001;
86:227 34.
[41] Meszaros I, Morocz J, Szlavi J, Schmidt J, Tornoci L,
Nagy L, et al. Epidemiology and clinicopathology of
aortic dissection. Chest 2000;117:1271 8.
[42] Nienaber CA, Eagle KA. Aortic dissection: new frontiers in diagnosis and management: Part II: therapeutic
management and follow-up. Circulation 2003;108:
772 8.
[43] Nienaber CA, Eagle KA. Aortic dissection: new frontiers in diagnosis and management: Part I: from etiology to diagnostic strategies. Circulation 2003;108:
628 35.
[44] Hagan PG, Nienaber CA, Isselbacher EM, Bruckman D, Karavite DJ, Russman PL, et al. The International Registry of Acute Aortic Dissection (IRAD):
new insights into an old disease. JAMA 2000;283:
897 903.
[45] Mehta RH, OGara PT, Bossone E, Nienaber CA, Myrmel T, Cooper JV, et al. Acute type A aortic dissection
in the elderly: clinical characteristics, management, and
outcomes in the current era. J Am Coll Cardiol 2002;
40:685 92.
[46] Kaji S, Nishigami K, Akasaka T, Hozumi T, Takagi T,
Kawamoto T, et al. Prediction of progression or regression of type A aortic intramural hematoma by computed tomography. Circulation 1999;100(Suppl 19):
II281 6.
[47] Nienaber CA, von Kodolitsch Y, Nicolas V, Siglow V,
Piepho A, Brockhoff C, et al. The diagnosis of thoracic
aortic dissection by noninvasive imaging procedures.
N Engl J Med 1993;328:1 9.
[48] LePage MA, Quint LE, Sonnad SS, Deeb GM, Williams DM. Aortic dissection: CT features that distinguish true lumen from false lumen. AJR Am J
Roentgenol 2001;177:207 11.
[49] Ganaha F, Miller DC, Sugimoto K, Do YS, Minamiguchi H, Saito H, et al. Prognosis of aortic intramural
hematoma with and without penetrating atherosclerotic
ulcer: a clinical and radiological analysis. Circulation
2002;106:342 8.

[50] Nienaber CA, von Kodolitsch Y, Petersen B, Loose R,
Helmchen U, Haverich A, et al. Intramural hemorrhage
of the thoracic aorta: diagnostic and therapeutic implications. Circulation 1995;92:1465 72.
[51] Stanson AW, Kazmier FJ, Hollier LH, Edwards WD,
Pairolero PC, Sheedy PF, et al. Penetrating atherosclerotic ulcers of the thoracic aorta: natural history and
clinicopathologic correlations. Ann Vasc Surg 1986;1:
15 23.
[52] Sawhney NS, DeMaria AN, Blanchard DG. Aortic
intramural hematoma: an increasingly recognized and
potentially fatal entity. Chest 2001;120:1340 6.
[53] Shimizu H, Yoshino H, Udagawa H, Watanuki A,
Yano K, Ide H, et al. Prognosis of aortic intramural
hemorrhage compared with classic aortic dissection.
Am J Cardiol 2000;85:792 5.
[54] von Kodolitsch Y, Csosz SK, Koschyk DH, Schalwat I,
Loose R, Karck M, et al. Intramural hematoma of
the aorta: predictors of progression to dissection and
rupture. Circulation 2003;107:1158 63.
[55] Song JK, Kim HS, Song JM, Kang DH, Ha JW, Rim
SJ, et al. Outcomes of medically treated patients with
aortic intramural hematoma. Am J Med 2002;113:
181 7.
[56] Kaji S, Akasaka T, Horibata Y, Nishigami K, Shono H,
Katayama M, et al. Long-term prognosis of patients
with type a aortic intramural hematoma. Circulation
2002;106(12 Suppl 1):I248 52.
[57] Song JK, Kim HS, Kang DH, Lim TH, Song MG, Park
SW, et al. Different clinical features of aortic intramural hematoma versus dissection involving the ascending aorta. J Am Coll Cardiol 2001;37:1604 10.
[58] Coady MA, Rizzo JA, Hammond GL, Pierce JG, Kopf
GS, Elefteriades JA. Penetrating ulcer of the thoracic
aorta: what is it? How do we recognize it? How do we
manage it? J Vasc Surg 1998;27:1006 15 [discussion:
1015 6].
[59] Wann S, Jaff M, Dorros G, Sampson C. Intramural
hematoma of the aorta caused by a penetrating atheromatous ulcer. Clin Cardiol 1996;19:438 9.
[60] Harris JA, Bis KG, Glover JL, Bendick PJ, Shetty A,
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
585
Brown OW. Penetrating atherosclerotic ulcers of the

aorta. J Vasc Surg 1994;19:90 8 [discussion: 98 9].
Rubinowitz AN, Krinsky GA, Lee VS. Intramural hematoma of the ascending aorta secondary to descending thoracic aortic penetrating ulcer: findings in two
patients. J Comput Assist Tomogr 2002;26:613 6.
Quint LE, Williams DM, Francis IR, Monaghan HM,
Sonnad SS, Patel S, et al. Ulcerlike lesions of the aorta:
imaging features and natural history. Radiology 2001;
218:719 23.
Tittle SL, Lynch RJ, Cole PE, Singh HS, Rizzo JA,
Kopf GS, et al. Midterm follow-up of penetrating
ulcer and intramural hematoma of the aorta. J Thorac
Cardiovasc Surg 2002;123:1051 9.
von Kodolitsch Y, Nienaber CA. Ulcer of the thoracic
aorta: diagnosis, therapy and prognosis. Z Kardiol 1998;
87:917 27.
Matsunaga N, Hayashi K, Sakamoto I, Ogawa Y, Matsumoto T. Takayasu arteritis: protean radiologic manifestations and diagnosis. Radiographics 1997;17:
579 94.
Hunder G. Vasculitis: diagnosis and therapy. Am J
Med 1996;100:37S 45S.
VanDyke CW, White RD. Congenital abnormalities
of the thoracic aorta presenting in the adult. J Thorac
Imaging 1994;9:230 45.
Ouriel K, Greenberg RK. Endovascular treatment
of thoracic aortic aneurysms. J Card Surg 2003;18:
455 63.
Parker MS, Matheson TL, Rao AV, Sherbourne CD,
Jordan KG, Landay MJ, et al. Making the transition:
the role of helical CT in the evaluation of potentially
acute thoracic aortic injuries. AJR Am J Roentgenol
2001;176:1267 72.
Mirvis SE, Shanmuganathan K, Miller BH, White CS,
Turney SZ. Traumatic aortic injury: diagnosis with
contrast-enhanced thoracic CT five-year experience
at a major trauma center. Radiology 1996;200:413 22.
Ahrar K, Smith DC, Bansal RC, Razzouk A, Catalano
RD. Angiography in blunt thoracic aortic injury.
J Trauma 1997;42:665 9.
Radiol Clin N Am 42 (2004) 587 601
CT and MR imaging of pericardial disease

Rainer Rienmuller, MDa,*, Reinhard Groll, MDa, Martin J. Lipton, MDb
a
Division of General Diagnostic Radiology, Interdisciplinary Cardiac Imaging Centre, Medical University of Graz,
Auenbruggerplatz 9, Graz 8036, Austria
b
Department of Radiology, Brigham and Womens Hospital, Harvard Medical School, 75 Francis Street,
Boston, MA 02115,USA
Chest radiography and echocardiography are the

initial imaging techniques used to evaluate suspected
or known pericardial disease. Despite many advantages, echocardiographic methods still have limitations, including small and often limited field of view;
occasionally poor acoustic windows; problems related
to anatomic factors, such as severe emphysema, large
calcified plaques, and deep thoracic diameter; and
limited reproducibility, especially of quantitative measurements. Furthermore, it occasionally is difficult to
distinguish between parapericardial fat, fluid, and
fresh clots.
Advanced multislice and multidetector CT and
electron-beam tomography and advanced MR imaging technologies are able to image the entire heart
and the complete chest with high spatial resolution
to below 1 mm. Furthermore, with continuous shortening of the imaging time per slice or volume it has
become more and more feasible to study the function, even for organs moving as fast as the heart.
Cine imaging using CT or MR imaging technologies is no longer merely a dream but a practical
reality. A holistic approach with qualitative and quantitative morphologic and functional analyses becomes not only possible, but also necessary for
better understanding of the normal and pathologic
organ function.
For easier understanding of the visualized morphology and function, as seen by CT and MR imag-
E-mail address: rainer.rienmuellerl@meduni-graz.at
(R. Rienmuller).
ing, some basic anatomic and functional aspects of the

pericardium and its disease are briefly mentioned.
Anatomy
The pericardium consists of two layers: an outer
fibrous layer (the fibrous pericardium) and an inner
visceral layer (the epicardium) creating an inner sac,
the pericardial cavity [1]. The wall thickness of the
inner layer varies between 0.05 and 1 mm [2], being
thicker above (eg, right ventricular myocardium) and
thinner along the thicker myocardial wall of the left
ventricle [3]. The inner surface of the outer layer is
lined by a layer of mesothelial cells producing serous
fluid. Under physiologic conditions the pericardial
sac contains 20 to 25 mL of serous fluid, which may
vary considerably in different individuals [4].
Functional anatomy
Histologically, the outer layer consists of just two
and the inner layer of three superimposed network-like
connective tissue structures giving the outer layer
more elasticity and distensibility; this is in contrast to
the inner layer, which appears less distensible and
stiffer [3]. That way, with an increasing pericardial
effusion, which is accompanied by increasing intrapericardial pressure, the outer layer stretches and
expands outward to prevent tamponade until it reaches
maximal distensibility. Next, the rising intrapericardial
pressure is directed to the inner layer (with lower
distensibility) impeding the normal diastolic filling
of the right ventricle and later also of the left, leading
to increased filling pressures, decreased cardiac out-
doi:10.1016/j.rcl.2004.03.003
588
R. Rienmuller et al / Radiol Clin N Am 42 (2004) 587601
put, and eventually clinical tamponade unless pericardiocentesis is performed.
Topographic anatomy
The pericardial sac encloses the heart, the proximal part of the ascending aorta, the pulmonary trunk,
and short segments of the left pulmonary veins. The
pars diaphragmatica of the pericardial sac is anchored
to the central tendon of the diaphragm. The sternocostal components of the pericardial sac are anchored
by the pericardio phrenic ligament to the diaphragm.
Superiorly lies the sternopericardial ligament.
The posterior and both lateral areas of the pericardium are connected to the aorta, tracheal bifurcation, and the right and left mediastinal pleura by
connective tissue. The retrosternal space is filled by
various amounts of fatty tissue (plica adiposa), which
may also be found in the recessus costomediastinalis.
Upward (cranially) approximately 1 to 1.5 cm below
the origin of the brachycephalic trunk the pericardial
sac is reflected onto itself creating the superior
junction line around the ascending aorta and the
pulmonary trunk until it reaches the ligamenta arteriosum [2].
The second pericardial junction line runs nearly
vertically from the superior to the inferior caval veins
enclosing in part both vessels. Because of the variety
in number and location of the pulmonary veins entering the left atrium the course of this junction line is
variable. A part of the left atrium is covered, but only
by the outer pericardial layer creating with the left
arterial wall the so-called mesocardium [2].
The transversal sinus of the pericardium localized
between ascending aorta and pulmonary trunk and
between the superior caval vein and left atrium shows
a variable course and may, dependent on the amount
of pericardial fluid, show a number of recesses, just
Fig. 1. Transverse drawings of the pericardial sinuses and

recesses at three closely adjacent slice levels (A C) above
the heart base. AA, ascending aorta; B, bronchus intermedius; DA, descending aorta; E, esophagus; IAR, inferior
aortic recess; LA, left atrium; LAA, left atrial appendage;
LPR, left pulmonic recess; LPVR, left pulmonic vein recess;
LSPV, left superior pulmonary vein; MPA, main pulmonary
artery; OS, oblique sinus; PCR, postcaval sinus; PRP,
posterior pericardial recess; RA, right atrium; RAA, right
atrial appendage; RPA, right pulmonary artery; RPR, right
pulmonic recess; RPVR, right pulmonic vein recess; RSPV,
right superior pulmonary vein; RVAT, right ventricular
outflow tract; SAR, superior aortic recess; SVC, superior
vena cava.
recently described [5], which may be misinterpreted

as lymphadenopathy or other mediastinal diseases if
the interpreter is not aware of these anatomic variations (Fig. 1). The sinus obliquus of the pericardium
is created by a blind pouch confined laterally by the
pulmonary veins, cranially by the left atrium, and
posteriorly by the pars dorsalis pericardii [4].
Pericardial function
It seems appropriate to distinguish between anatomic and mechanical function of the pericardium.
The anatomic functions consist of the following:
Fixing of cardiac position in the thoracic cavity
[1,6]
Isolation of adjacent thoracic structures to re-
duce spread of pathologic processes to or from

the heart [1,6]
Reduction of the friction resistance between the
outer and inner pericardial layer by the presence
of pericardial fluid [6]
Inclusion of blood and lymphatic vessels and of
cardiac nerves [3]
The function of the pericardium is not entirely
understood. Certainly patients survive when it is partially absent, congenitally or acquired. Electron-beam
CT has been used to explore its function [7 9].
Mechanical function of the pericardium does
the following:
589
Visualization of the pericardium

In CT and MR imaging the normal pericardium is
visualized as a pencil-thin line in front and along the
right atrioventricular groove and the right ventricle
(separated from the right ventricular myocardium by
connective fatty tissue) as far as the apex of the left
ventricle and cranially until the superior junction line
[13,14]. Along the right heart the thickness of the
smooth pericardium should not exceed 1 to 1.5 mm.
Along the left ventricular wall the pericardial line is
even thinner and the amount of subepicardial connective fatty tissue even narrower than along the right
ventricle. The pericardium frequently may not be
visualized along the left myocardial wall, but it is
visualized in front of the left atrioventricular groove
because of the (visual) presence of connective fatty
tissue (Fig. 2). In case of nonvisualized pericardium
along the left ventricle the total or partial lack of the
pericardium may be excluded if the heart shows
normal mediastinal position (no shifting to the left
chest space) and no regional bulging of the left-sided
cardiac structures [15].
MR imaging seems generally superior to CT for
visualization of cardiac structures because of its better
contrast resolution and direct multiplanar imaging
capabilities. In daily practice, however, studying patients with suspected or known constrictive pericarditis by MR imaging (without CT) may be difficult.
Prevents the (acute) dilatation of the cardiac

chambers [6] and is probably thicker along

the right than along the left ventricular myocardium [6]
In case of increased systemic circulatory resistance, the right ventricular stroke volume is
adapted to the reduced left ventricular stroke
volume [6]
Prevents ventriculoatrial reflux if the enddiastolic pressures are elevated [6,10]
Supports atrial filling by creating negative pericardial pressure during systole
Contributes, in experimental studies, to prevent
myocardial hypertrophy in long-term physical
stress [10,11]
Contributes to the regulation of the venous pressure [11] (there is no increase of the end-diastolic
right ventricular pressure without pericardium)
Contributes reflectory to the regulation of blood
pressure and of the heart rate [11,12]
Fig. 2. Electron-beam tomography image of normal pericardial line (arrow) seen in front of the right atrium, right
ventricle, and anterior to the left ventricular apex. Note how
it is separated from the heart by periepicardial fat and connective tissue. (From Groll R, Schaffer GJ, Rienmuller R.
Pericardial sinuses and recesses: findings at electrocardiographically triggered electron-beam CT. Radiology 1999;
212:69 73; with permission.)
590
Table 1
Imaging methods: heart, anatomy, function
Pericardium
Epicardium
Myocardium
Valves
Cavity
Coronary wall
Coronary lumen
RV function
LV function
Myocardial perfusion
Coronary flow
Intracavitary flow
CT
EBT
MR imaging
++
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+++
+++
++
(+ + +)
(+ + +)
+++
+++
++
(+ + +)
+++
++
+
+
+
+
+
++
+
+
++
+
imaging signals may be helpful in elucidating these

findings (Fig. 4).
The sequels of any pericarditis may be:
Restitutio ad integrum [16]
Exudate transformation to hyaline scarred tissue
Adhesion of both pericardial layers scarred as
Chronic-adhesive pericarditis of both visceral and
parietal layers
Chronic-constrictive fibrous or calcific pericarditis
Chronic-adhesive mediastinopericarditis as accretio cordis [16]
, not possible; +, possible; + +, more suitable; + + +, best

suitable; (+ + +), work in progress.
Abbreviations: EBT, electron-beam tomography; LV, left
ventricle; RV, right ventricle.
This is because it may not be possible to identify the

pericardial line, especially if thin, from signal-free
areas of calcification, fibrosis, fluid, or air. Myocardial calcification may be very difficult to identify because if positive it suggests the presence of chronic
or subacute perimyocarditis. This is important because it suggests possible regional or global myocardial fibrosis, a contraindication for pericardiectomy.
Table 1 provides an assessment of CT, electron-beam
tomography, and MR imaging for evaluating cardiac
morphology and function with current state-of-theart equipment.
Pericardial pathology
Pericarditis
Any pericardial stimulus of a certain threshold
may cause a classic inflammatory reaction [10], the
extent and distribution of which may be seen on CT
or MR imaging as a thickened pericardial line ( 1
2 mm). As long as this pericardial line appears
smooth from both sides of the pericardial sac, this
finding may be regarded as acute; if it appears
irregularly thickened it is regarded as a chronic
inflammatory or postinflammatory process (Fig. 3)
[13]. This inflammatory reaction is usually accompanied by a different amount of exudation into the
pericardial cavity [10]. The exudate may be serous,
fibrinous, purulent, hemorrhagic, or mixed and a
careful analysis of measured CT values or MR
Fig. 3. Electron-beam tomography slices (A) above the base

of the heart with replacement of the aortic valve (B) below
the middle of the left ventricle with irregularly thickened pericardium around the heart. Because both caval veins (arrows)
and the partly visualized pulmonary veins show no enlargement and the thickened periepicardial line is separated
from cardiac structures by subepicardial fat and connective
tissue, pericardial constriction may be excluded because
there are no morphologic determinants of inflow obstruction.
591
Pericardial effusion
Any increase of pericardial fluid, in dependence
on the distensibility characteristics of the individual
patients pericardium, increases the intrapericardial
pressure. Generally, slow increase of pericardial fluid
up to 3 L is well tolerated [17]. The rapid development with a sudden increase of pericardial fluid,
however, may cause sudden tamponade without prodromal symptoms (Fig. 5). This means that it is not
the volume of pericardial fluid but rather the intrapericardial pressure that is the most important factor
in determining the risk of tamponade. This intra-
Fig. 4. Coronal (A), sagittal (B), and transversal (C) MR

images showing a pericardial effusion 2.5 cm wide encircling the whole heart from the diaphragm to the superior
junction line.
In connection with special pathologic conditions,

hydropericardium, hematopericardium, chylopericardium, or pneumopericardium may be seen. CT values
and MR imaging signal analysis may be helpful for
differential diagnostic purposes.
Fig. 5. (A) Drawing of midventricular-level slice of pericardial effusion illustrating the subepicardial fat and connective tissue and normal configuration of both ventricles
and both atria. (B) Drawing of midventricular-level slice of
pericardial effusion with no visualization of the subepicardial fat and connective tissue, with tubelike configuration of
both ventricles because of elevated intrapericardial pressure, causing compression of both ventricles and enlargement of both atrial auricles, which are seen near the cardial
apex, because of the presence of morphologic (signs) determinants of pericardial tamponade. (From Rienmueller R,
Seiderer M, Doliva R, Kemkes B, Lissner J. Pericardial and
congestive heart failure: diagnostic with CT- and MR-imaging. Ann Radiol 1986;29:95 100; with permission.)
592
pericardial pressure is very much related to individual

pericardial distensibility (eg, it is decreased in case of
scarring) and to the acute extent of the distention of
the pericardial layers.
Sequels of intrapericardial pressure increase

Increase of intrapericardial pressure may result in
the following:
Restriction of atrial filling [10]
Restriction of diastolic ventricular filling
be seen with electron-beam CT and in cine MR

imaging studies as in echocardiographic studies.
Localized intrapericardial hematomas, frequently
found after cardiac surgery in front of the right (and
very rarely in front of the left) heart compressing
the right ventricle or right atrium, may simulate
clinical symptoms of cardiac tamponade. For diagnostic purposes by CT or MR imaging it is easier than
by echocardiography to prove or exclude right-sided
effusive constrictive periepicarditis, the first being an
indication for pericardiectomy and the second for
a pericardiocentesis.
[18 20]
Decrease of end-diastolic volume [18,21]
Increase of end-diastolic pressures in all cardiac
chambers [10,22,23]
Constrictive pericarditis
Definition
Narrowing of the right atrium with venous blood

inflow obstruction [19]

Decrease of stroke volume [10,21]
Decrease of arterial blood pressure [10,19,21]
Decrease of blood pressure amplitude [10]
Decrease of coronary blood flow [10]
Reduction of cardiac output [19,21]
Compensatory mechanisms to reduce the sequels

of intrapericardial pressure increase
The following mechanisms may become activated
to compensate the negative sequelae of intrapericardial pressure elevation:

Peripheral vasoconstriction [21]

Tachycardia [10]
Higher oxygen use [10]
Activation of the cardiac sympathetic nervous
system accompanied by increase of catecholamine levels [24]
Increase of the venous pressure [10,22]
The term constrictive pericarditis [26,27] is applied to a disorder in which inflammatory or noninflammatory pericardial processes have caused
scarring or calcification of one or both pericardial
layers, which leads to constriction [1] and frequently
to compression of the underlying cardiac chambers.
As a sequel of these pathologic pericardial changes,
the normal physiologic compliance of one or both
pericardial layers is lost resulting mechanically in
impaired (restrictive) filling of the cardiac chambers
during diastole.
Clinical symptoms of pericardial constriction

The clinical symptoms of cardiac failure in patients with pericardial constriction are variable and
numerous [28 31]. Depending on the hemodynamic
effects, the constriction may be mild, moderate,
severe [1], or occult (masked) [32,33]. The symptoms
of heart failure may be even more variable when the
pericardial constriction is associated with additional
myocardial, valvular (frequent), coronary, or pulmonary disease [34].
Pericardial tamponade
If the intrapericardial pressure is elevated above
a value of 20 mm Hg, no effective filling of the
ventricles is possible resulting in pericardial tamponade [24]. In CT and MR imaging with pericardial
effusion the loss of definition of the subepicardial
space (connective and fatty tissue) with compression
of the ventricles and deformation of the atria (with the
atrial auricles being seen to the level of the cardiac)
suggests the presence of pericardial tamponade
(Fig. 5) [25]. The swinging motion of the heart may
Mechanical sequels of cardiac compression

As a result of the cardiac compression by scarred
or calcified pericardium the following mechanical
changes may be observed:
Restrictive filling of the ventricles [35]
Restriction of diastolic ventricular dilatation
[19,27]
Decrease of ventricular volumes [27,36]

Decrease of stroke volume [27,35,37]

Decrease (minimal) of cardiac output [19,37]
Decrease of ejection fraction [6,27,38]
Hepatosplenomegaly, ascites [19]
Atrial fibrillation [37]
593
cause the wall is thin, usually not measuring more

than 5 to 7 mm. In pericardial constriction the following are controversial issues:
Duration of conservative therapy for symptoms
(diuretics, glycosides, and so forth) [31,34,46]

Hemodynamic sequels of cardiac compression
Hemodynamically, the cardiac compression by
scarred or calcified pericardium leads to the following:
Increase of diastolic pressures and their equili-

bration in the right atrium, right ventricle, left

atrium, left ventricle, pulmonary trunk [27], and
pulmonary wedge pressure [22]
Pulmonary and systemic venous congestion
[19,37]
Increase of venous blood pressure [35]
Systemic blood pressure remains in normal
range with tendency to decrease [19,35]
Heart rate remains constant with a tendency to
increase during stress [19]
Prolongation of circulation time [39]
Increase of circulatory blood volume [39]
Progressive restriction of ventricular filling with
exaggerated breathing and variation of the
stroke volume (pulsus paradoxus) [15].
Therapy for fibrotic or calcified constrictive

pericarditis
Because the chronic fibrosis or calcified pericardial changes are irreversible, the only treatment that
can improve or normalize the restrictive cardiac hemodynamics, thereby allowing better filling of the
cardiac chambers, is total or partial pericardiectomy
[30,31,40 43]. In reality complete pericardiectomy is
surgically not possible and frequently the term pericardial fenestration is used.
Depending on the type of pericardial constriction
(global, right-sided, left-sided, annular) [43] the following thoracotomy approaches are applied: leftsided, bilateral, or transsternal bilateral thoracotomy
[44] or median sternotomy [41,43]. For pericardiectomy a scalpel, harmonic scalpel [45], electrocautery,
or simply a manual surgical approach may be used. At
the end of a pericardial fenestration procedure measurements of intracardiac end-diastolic pressures are
recommended to ensure that all constrictive regions of
the pericardium were indeed resected. When performing epicardiotomy (inner layer) there is a high
risk of right ventricular myocardial perforation be-
Optimal timing of surgery

Best method of thoracotomy [47]
Suitable extent and optimal region of periepi-
cardiectomy (neither the outer nor the inner

layer of the pericardium can be completely dissected; only outer or inner layer fenestrations of
various extent are possible) [48]
Selection of patients who benefit from surgical
treatment [47]
The reported intraoperative and perioperative
mortality rate in patients with pericardial constriction (V 50%) [40,41,49,50]
A review of the literature indicates that the presence of the myocardial factor (myocardial atrophy or
fibrosis) may lead to an acute dilatation of one or
both cardiac chambers or to severe cardiac insufficiency with dilatation of one or both ventricles
following surgery [1,50 54].
CT and MR imaging diagnostic criteria of

pericardial constriction
For diagnostic morphologic evaluation of patients
with known or suspected pericardial constriction
using CT or MR imaging it is absolutely necessary
to image the complete heart from aortic arch to the
diaphragm. In CT it is recommended to perform this
study with and without contrast agent to display the
cardiac cavity and frequently this allows the differentiation of calcified plaque from contrast-enhanced
structures (Fig. 6).
During the early (arterial) phase of intravenous
contrast agent enhancement of the pericardium in
acute (nonfibrotic) pericarditis occurs in contrast to
chronic (fibrotic) pericarditis. For MR imaging contrast enhancement is necessary if a tissue perfusion
related question needs resolving, such as the distinction between chronic (fibrotic) pericarditis, which
enhances late, and acute (nonfibrotic) pericarditis
without late enhancement.
As shown in Table 1, a careful systematic analysis
of the periepicardium includes the size and configuration of the cardiac chambers, the superior and
inferior vena cavae, the pulmonary trunk (as compared with the ascending or descending thoracic
aorta), and the atrio-ventricular grooves, permitting
594
Fig. 6. Electron-beam tomography slices at the level of the superior (A) and inferior (B) caval veins, above (C) and through the
midventricular level (D) of the heart, showing enlargement of both caval veins (coronary sinus), of the atria, and normal
configuration of both ventricles with calcified pericardium from the pars diaphragmatica (B) pericardii around the heart (C,D) until
the superior junction line, confirming global type of calcified pericardial constriction. Only at the left ventricular apex is the
pericardium not calcified. (Calcified pericardium anterior to the apex of the left ventricle may be seen without pericardial
constriction). The intramyocardial calcification arising near the calcified pericardium dorsal of the left ventricle (D) is suggestive of
previous perimyocarditis. The posterolateral wall of the left ventricle and the interventricular septum (systolic image) have a normal
wall thickness, which excludes myocardial atrophy. At pericardiectomy caution is necessary during pericardial decortication to
prevent myocardial damage. Intramyocardial calcification may be difficult to see and overlooked if only MR imaging is used.
a diagnosis of pericardial constriction. If all these

morphologic determinants of pericardial constriction
are present, the diagnosis of pericardial constriction is
absolutely certain (accuracy 100%) [43].
Morphologic types of pericardial constriction

Both CT and MR imaging [43] can identify and
distinguish between the following types of pericardial
constriction (Box 1):
Global: bilateral thickening or calcification of the
pericardium along both ventricles and enlargement of both atria, superior vena cava, and
inferior vena cava (Fig. 7)
Annular: bilateral thickening or calcification of

the pericardium primarily in the atrioventricular grooves with narrowing of both grooves;
normal-sized ventricles; and enlargement of
both atria, superior vena cava, and inferior
vena cava (Fig. 8) [43,55,56,66,67]
Left-sided thickening or calcification of the
pericardium along the compressed left ventricle with narrowing of the left atrioventricular
groove; intraventricular septum straight or bent
toward the left; enlargement of both atria, superior vena cava, and inferior vena cava; and
normal-sized right ventricle (Fig. 9A)
Right-sided thickening or calcification of the pericardium anterior to the compressed right ventricle with the interventricular septum straight
595
or bent to the right; narrowing of the right

atrioventricular groove; and enlargement of the
right atrium, superior vena cava, and inferior
vena cava (Fig. 9B)
Epicardial: the global or focal form of constriction
is predominantly caused by the involvement of
the epicardial layer [43]
Effusive: the configuration of the epicardium does
not change regardless of the amount of
pericardial fluid, but there is general epicardial
constriction and pericardial effusion [43,57]
CT and MR imaging parameters of myocardial

atrophy or fibrosis in pericardial constriction
Previous studies [43,51,52,54] have shown that
unrecognized myocardial atrophy or fibrosis (myocardial factor) is the most frequent cause of the high
intraoperative and perioperative mortality in patients
referred for pericardiectomy, with constrictive pericardial disease. In a retrospective study [43] it was
also shown that using CT or MR imaging (Table 2,
Fig. 10) the perioperative mortality rate can be
reduced by excluding patients with myocardial atrophy or fibrosis from pericardiectomy (Fig. 11).
Box 1. Morphologic signs (determinants)

of pericardial constriction
Global or focal periepicardial thickening, calcification, or both
Tube-like configuration of one or
both ventricles
Narrowing of one or both atrioventricular grooves
Bent or sinuous appearance of the interventricular septum
Enlargement of one or both atria
Enlargement of the diameters of the
superior and inferior vena cava
in comparison with the diameters
of the descending aorta (1 1 and
2 1, respectively)
From Rienmu
ller R, Gu
rgan M, Erdmann
E, Kemkes BM, Kreutzer E, Weinhold CH.
CT and MR evaluation of pericardial constriction: a new diagnostic and therapeutic concept. J Thorac Imaging 1993;8:
108 21; with permission.
Fig. 7. (A) Drawing of midventricular-level slice in the

global form of pericardial constriction, with tubelike (compressed) configuration of both ventricles, fibrous or calcified
thickening of the periepicardium along both ventricles, and
narrowing of the atrioventricular grooves. Notice that the
thickened periepicardium is separated from the right ventricular wall by a thin layer of subepicardial fat and connective tissue. (B) Drawing of midventricular-level slice in
restrictive cardiomyopathy with enlarged but normal configuration of the atria, normal-sized ventricles, and no
thickening of the periepicardium. The periepicardium
usually is not visible on CT studies of patients with restrictive cardiomyopathy. (From Rienmuller R, Gurgan M,
Erdmann E, Kemkes BM, Kreutzer E, Weinhold CH. CT
and MR evaluation of pericardial constriction: a new diagnostic and therapeutic concept. J Thorac Imaging 1993;8:
108 21; with permission.)
The etiology of myocardial atrophy and fibrosis

remains controversial [51,52,54]. Myocardial fibrosis
is thought to be a result of previous chronic and
severe perimyocarditis and the atrophy is a sequel of
chronically reduced ventricular work caused by pericardial constriction. Surgery is frequently recommended as soon as possible.
In the authors study [43] all patients with leftventricular myocardial atrophy who expired at sur-
596
appears irregular then chronic pericarditis is present.

Because there is no direct relationship between the
extent of thickness of the pericardial layers and its
hemodynamic effectiveness pericardial constriction
can be diagnosed only in the presence of all the
remaining morphologic determinants of constriction
as described in Table 1 [43].
If pericardium seems to be only minimally thickened [58 60] around the heart and all the remaining morphologic determinants of constriction are
seen, pericardial constriction is present with smooth
Fig. 8. Drawing of midventricular-level slice of the annular

form of periepicardial constriction, with thickening of the
periepicardium mainly in front of the atrioventricular grooves
and narrowing of both grooves. The atria are enlarged, and
the ventricles are of normal size. (From Rienmuller R,
Gurgan M, Erdmann E, Kemkes BM, Kreutzer E, Weinhold
CH. CT and MR evaluation of pericardial constriction: a
new diagnostic and therapeutic concept. J Thorac Imaging
1993;8:108 21; with permission.)
gery revealed a history of severe chronic coronary

artery disease and bypass surgery, suggesting that
chronic ischemia may contribute to or be the cause of
myocardial atrophy and fibrosis. The excellent soft
tissue contrast resolution of MR imaging is especially
suitable for excluding right ventricular myocardial
fibrosis by demonstrating the presence of subepicardial fat between the inner pericardial layer and the
right ventricular myocardial wall (Fig. 12).
In the future the use of fast CT and MR imaging
techniques will enable more accurate and precise
measurement of the left and right myocardial wall
and its dynamic changes in thickening during the
cardiac cycle. Together with myocardial (and pericardial) perfusion measurements this will allow more
improved diagnosis of myocardial atrophy and fibrosis, the presence of which contraindicates pericardiectomy, because of the risk of acute intraoperative or
postoperative cardiac dilatation or rupture.
Possible pitfalls in CT or MR imaging diagnostic

interpretation of pericardial constriction
Pericardium may appear to be only minimally and
partially thickened [58 60], whereas all other morphologic determinants of pericardial constriction are
present. It must be emphasized that any thickening of
the pericardial line more than a pencil line suggests
the presence of acute pericarditis if both outer and
inner pericardial layers appear smooth. If the line
Fig. 9. (A) Drawing of midventricular-level slice of the leftsided form of pericardial constriction. The thickened
periepicardium is separated by subepicardial fat from the
compressed left ventricle. The interventricular septum often
is bent to the left. (B) Drawing of midventricular-level slice
of the right-sided form of pericardial constriction. The
thickened periepicardium is separated by subepicardial fat
from the compressed right ventricle. The interventricular
septum often is bent to the right. (From Rienmuller R,
Gurgan M, Erdmann E, Kemkes BM, Kreutzer E, Weinhold
CH. CT and MR evaluation of pericardial constriction: a
new diagnostic and therapeutic concept. J Thorac Imaging
1993;8:108 21; with permission.)

Table 2
CT and MR imaging criteria of myocardial atrophy
or fibrosis
Left ventricle
Right ventricle
Interventricular
septum < 1 cm
Periepicardium not
separated from
ventricular wall
Irregular wall thickening
Ventricular wall < 1 cm

Systolic diastolic wall
thickness changes < 40%
LVMM/EDV ratio < 1.1
Abbreviations: EDN, end-diostotic volume; LVMM, left

ventricular muscle mass.
From Rienmuller R, Gurgan M, Erdmann E, Kemkes BM,
Kreutzer E, Weinhold CH. CT and MR evaluation of pericardial constriction: a new diagnostic and therapeutic concept. J Thorac Imaging 1993;8:108 21; with permission.
shrinkage of the pericardial layers (causing diastolic

intracavitary inflow obstruction). In a recent pathoanatomic study [58] it was reported that in pericardial
constriction the pericardial thickness ranged from 1 to
7 mm (mean, 4 mm). The same working group [61]
reported patients with constrictive pericarditis and
pericardial thickness less than or equal to 2 mm, and
histopathologic abnormalities, such as mild and focal
inflammation, including fibrosis, calcification, fibrin
deposition, and focal noncaseiform granulomas.
There are reports of patients [60] diagnosed as
having acute pericardial constriction by echocardiographic and cardiac catheter-methods (without CT or
MR imaging) in whom, by just conservative treatment (without pericardiectomy), the symptoms of
pericardial constriction were found to regress to
normal hemodynamics (at least at rest). These reports
may be supported by the following observations. In
patients with acute tuberculous pericarditis the authors found various amounts of pericardial fluid and
both outer and inner layers of the pericardium were
thickened and these enhanced after intravenous contrast agent administration. Simultaneously, all of the
previously mentioned morphologic determinants
listed in Box 1 of intracavitary blood flow obstruction
confirmed the diagnosis of pericardial constriction.
During the tuberculostatic treatment, antiphlogistic
drugs were additionally prescribed and in a period
of about 1 year the morphologic determinants of
intracavitary blood flow obstruction disappeared (at
least at rest).
It may be hypothesized that even in the stage of
acute tuberculous pericarditis with pericardial effusion the inner and to a lesser degree the outer layer of
the pericardium may shrink, whereby the process of
shrinkage may be at least arrested or even regress
597
because of the antituberculosis treatment. To select

the optimal treatment (surgery or conservative) it is
necessary to distinguish between acute, transient
(nonfibrotic) [58,60], and chronic (fibrotic) constrictive pericarditis. Contrast agents should be applied in
CT and MR imaging studies to distinguish these
disease entities.
Postpericardiectomy hemodynamic and clinical

results
In terms of absolute measurements of cardiac
hemodynamics, the number of reports concerning
Fig. 10. (A) Drawing of midventricular-level slice in leftsided myocardial atrophy with thinning of the posterolateral wall and the interventricular septum. (B) Drawing of
midventricular-level slice of the right myocardial fibrosis
showing irregular thickening of the right ventricular wall and
focal nonseparable thickened periepicardium. (From Rienmuller R, Gurgan M, Erdmann E, Kemkes BM, Kreutzer E,
Weinhold CH. CT and MR evaluation of pericardial constriction: a new diagnostic and therapeutic concept. J Thorac
Imaging 1993;8:108 21; with permission.)
598
Fig. 11. Perioperative mortality (percentage) in patients with pericardial constriction from retrospective (1980 1984) and
prospective (1985 1991) CT and MR imaging studies. In a retrospective study 6 of 20 patients expired with the previously
described determinants of myocardial atrophy and fibrosis, respectively (mortality rate could be decreased from 30% 16.5%). In
the prospective study 5 of 30 patients expired, 3 because of the presence of myocardial atrophy-fibrosis (mortality 10%) and
2 (without myocardial atrophy-fibrosis) because of surgical complications (mortality 6.6%). (From Rienmuller R, Gurgan M,
Erdmann E, Kemkes BM, Kreutzer E, Weinhold CH. CT and MR evaluation of pericardial constriction: a new diagnostic and
therapeutic concept. J Thorac Imaging 1993;8:108 21; with permission.)
functional outcome and short- and long-term results in

patients who have undergone pericardiectomy was
until recently limited. A few studies reported improvement of clinical symptoms of cardiac insufficiency
from New York Heart Association classification stage
3 to 2 or from stage 2 to 1 [62,63]. Complete relief of
symptoms in survivors was reported to be about 50%
[42]. There is only limited information concerning the
frequency of recurrence following pericardiectomy.
What should be known before pericardiectomy

The present understanding of the morphology and
function of the pericardium [55] and of constrictive
pericardial disease, a review of the literature [30,31,
33,40,47,49,50,63 65] and of the authors patients
medical records, together with discussions with cardiologists and cardiac surgeons showed that the
following questions are most relevant for patients
with suspected or known pericardial constriction:
Is there pericardial constriction or does the pa

tient have a restrictive cardiomyopathy?

Is there myocardial fibrosis or atrophy?
What form of pericardial constriction is present?
Which type of thoracotomy should be used?
Where should pericardiectomy be done and
how much pericardial fenestration is possible
and necessary?
Will the patient benefit more from conservative
treatment or from pericardiectomy?

What is the most probable etiology of the
disease?
Summary
In patients with restrictive or constrictive cardiac
hemodynamics, in whom there is elevation of diastolic pressure in all four cardiac chambers, CT or MR
imaging can determine the presence or absence of the
morphologic determinants of pericardial constriction
to identify and characterize patients with pericardial
constriction. Diagnostic thoracotomy to distinguish
between pericardial constriction and restrictive cardiomyopathy is now considered obsolete [43].
Myocardial atrophy and fibrosis may be detected
preoperatively by CT or MR imaging. In these
patients pericardiectomy is contraindicated and cardiac transplantation should be considered as an alternative surgical treatment [43].
Presurgical planning is critical to determine the
form and extent of pericardial constriction because
this dictates the optimal thoracotomy approach. The
extent of disease and the area of periepicardial
fenestration, and the optimal sequence for performing
the periepicardial decortication (first along the left
ventricle, then in the left atrioventricular groove,
anterior to the pulmonary trunk, anterior to the right
599
Fig. 12. MR imaging slices (A) coronal, (B) through the midventricular level in the short axis view in a gradient echo white blood
pool image, (C) at midventricular level in long axis view, and (D) spin echo (black blood pool image) technique. The superior
caval vein (A) and the right atrium (A,C) are enlarged and the right ventricle is compressed (B D) by pericardial mass of
inhomogeneous signal intensity (B D). This pericardial mass is surrounded by calcified outer and inner layer of the pericardium
(difficult to see without CT). The subepicardial fat and connective tissue in high (white) signal intensity (D) between the inner
pericardial layer and the normal thickness of the ventricular myocardium exclude right myocardial fibrosis and atrophy.
ventricle, and finally in the area of the right atrioventricular groove in patients with global pericardial
constriction), can all be guided by imaging [43].
The continuous improvement of coronary artery
imaging using advance CT and MR imaging technology will in the near future replace coronary angiography in this disease entity. Cine CT and cine
MR imaging techniques will depict coronary artery
anomalies and interventricular septal motion [65],
and measure blood flow for the early recognition
of restrictive flows in subacute or masked pericardial constriction.
Intravenous contrast-enhanced CT or MR imaging
will in the future be increasingly applied for the differential diagnosis of acute (nonfibrotic) and chronic
(fibrotic) pericarditis and will play an important role
in triaging patients for either conservative therapy
or surgical pericardiectomy [43]. In addition, these
cross-sectional techniques will also be important for

identifying other diseases that mimic pericarditis and
pericardial constriction.
Acknowledgments
The authors are grateful to Mrs. Gollowitsch for
her secretarial assistance and Dr. U. Reiter for graphical preparation.
References
[1] Shabetai R. The pericardium. New York: Grune &
Stratton; 1981.
[2] Bargman W, Doerr W. Das herz des menschen. Stuttgart: Georg Thieme Verlag; 1963.
600
[3] Schorn J. Pathologie der Herzhullen. In: Bargmann W,

Doerr W, editors. Das herz des menschen. Stuttgart:
Georg Thieme Verlag; 1963. p. 865 93.
[4] Moncada R, Baker M, Salinas M, et al. Diagnostic role
of computed tomography in pericardial heart disease:
congenital defects, thickening, neoplasms and effusion.
Am Heart J 1982;103:263 82.
[5] Groll R, Schaffer GJ, Rienmuller R. Pericardial sinuses
and recesses: findings at electro cardiographically triggered electron-beam CT. Radiology 1999;212:69 73.
[6] Holt JP. The normal pericardium. J Cardiol 1970;26:
455 65.
[7] Smiseth OA, Refsum H, Junemann M, Sievers RE,
Lipton MJ, Carlsson E, et al. Ventricular diastolic
pressure-volume shifts during acute ischemic left ventricular failure in dogs. J Am Coll Cardiol 1984;3:
956 77.
[8] Junemann M, Smiseth OA, Refsum H, Sievers R, Lipton MJ, Carlsson E, et al. Quantification of effect of
pericardium on LV diastolic PV relation in dogs. Am J
Physiol 1987;21:H963 8.
[9] Smiseth OA, Frais MA, Junemann M, Kingma I,
Refsum H, Smith ER, et al. Left and right ventricular
diastolic function during acute pericardial tamponade.
Clin Physiol 1991;11:61 71.
[10] Schollmeyer P. Perikarditis. In: Reindell H, Roskam H,
editors. Erkrankungen des herzens. Berlin: Springer
Verlag; 1977. p. 665 75.
[11] Hort W. Der herzbeutel und seine bedeutung fur das
herz. In: Heilmeyer L, Prader A, Muller AF, Schoen R,
editors. Ergebnisse d inn medizin 1974;29:1 50.
[12] Tyberg THI, Goddyer AVN, Langou RA. Genesis of
pericardial knock in constrictive pericarditis. Am J
Cardiol 1980;46:570 5.
[13] Doppman JL, Rienmuller R, Lissner J, et al. Computed
tomography in constrictive pericardial disease. J Comput Assist Tomogr 1981;5:1 11.
[14] Silverman PM, Harell GS. Computed tomography of
the normal pericardium. Invest Radiol 1983;18:141 4.
[15] Breen JF. Imaging of the pericardium. J Thorac Imaging 2001;16:47 54.
[16] Bankl H. Pathologie der kreislauforgane. In: Holzner JH, editor. Arbeitsbuch der pathologie. Spezielle
pathologie. Munchen: Urban & Schwarzenberg Verlag;
1981. p. band II.
[17] Shabetai R, Meaney E. Haemodynamics of cardiac
restriction and tamponade. Am Heart J 1979;98:
780 5.
[18] Autenrieth G. Der tamponierende perikardergu. Internist 1980;21:17 24.
[19] Friedberg CK. Pericarditis adhaesiva: pericarditis constrictiva chronica. In: Friedberg Ch K, editor. Erkrankungen des herzens. Stuttgart: Georg Thieme Verlag;
1972. p. 994 1011.
[20] Isaacs JP, Berglund E, Sarnoff SJ. Ventricular function.
Am Heart J 1954;66 76.
[21] Isner JM, Carter BL, Bankoff MS, Kontam MA, Salem
DN. Computed tomography in the diagnosis of pericardial heart disease. Ann Intern Med 1982;97:473 9.
[22] Shabetai R, Fowler NO, Genton JC. Restrictive cardiac

disease: pericarditis and the myocardiopathies. Am
Heart J 1965;69:271 80.
[23] Lackner K. Clinical applications of CTT for evaluation
of ischemic heart disease: comparison with other imaging methods. In: Higgins CB, editor. CT of the heart
and the great vessels: experimental evaluation and clinical application. Mount Kisco, NY: Futura Publishing;
1983. p. 267 88.
[24] Pegram BL, Kardion MB, Bishop VS. Changes in left
ventricular internal diameter with increasing pericardial pressure. Cardiovasc Res 1975;9:707 14.
[25] Rienmueller R, Seiderer M, Doliva R, Kemkes B, Lissner J. Pericardial and congestive heart failure: diagnostic with CT- and MR-imaging. Ann Radiol 1986;29:
95 100.
[26] Sommerville W. Constrictive pericarditis: with special reference to the change in natural history brought
about by surgical intervention. Circulation 1968;
37,38(Suppl 5):102 22.
[27] Hirschmann JV. Pericardial constriction. Am Heart J
1978;96:110 22.
[28] Glockner JF. Imaging of pericardial disease. Magn
Reson Imaging Clin N Am 2003;11:149 62.
[29] Bubenheimer P, Villanyi J, Tollenaere P. Pericarditis
constrictiva als spatkomplikation von herzoperationen.
Z Kardiol 1985;74:91 8.
[30] Nishimura RA, Connoly DC, Parkin TW, Stanson AW.
Constrictive pericarditis: assessment of current diagnostic procedures. Mayo Clin Proc 1985;60:397 401.
[31] Tuna IC, Danielson GK. Surgical management of pericardial diseases. Cardiol Clin 1990;8:683 96.
[32] Bush CA. Occult constrictive pericarditis. In: Reddy
PS, Leon DF, Shaver JA, editors. Pericardial disease.
New York: Raven Press; 1982.
[33] Killian DM, Furiasse JG, Scanalon PJ, Loeb HS, Sullivan HJ. Constrictive pericarditis after cardiac surgery.
Am Heart J 1989;118:563 8.
[34] Lorell BH, Grossmann W. Profiles in constrictive pericarditis, restrictive cardiomyopathy, and cardiac tamponade. In: Grossmann W, editor. Cardiac catheterisation
and angiography. 3rd edition. Philadelphia: Lea &
Febiger; 1986. p. 427 45.
[35] Gaasch WH, Peterson KL, Shabetai R. Left ventricular
function in chronic constrictive pericarditis. Am J
Radiol 1974;34:107 10.
[36] Reedy PS, Curtiss EI, OToole JD, Shaver JA. Cardiac
tamponade: hemodynamic observations in man. Circulation 1978;58:265 72.
[37] Shabetai R, Fowler NO, Guntheroth WG. The hemodynamics of cardiac tamponade and constrictive pericarditis. Am J Cardiol 1970;26:480 9.
[38] Vogel JHK, Horgan JA, Strahl CL. Left ventricular
dysfunction in chronic constrictive pericarditis. Chest
1971;59:484 92.
[39] Stewart HJ, Heuer GJ. Chronic constrictive pericarditis. Arch Intern Med 1939;63:504 30.
[40] Culliford AT, Lipton M, Spencer FC. Operation for
chronic constrictive pericarditis: do the surgical ap-
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
proach and degree of pericardial resection influence the

outcome significantly? Ann Thorac Surg 1980;29:
146 52.
Yetkin U, Kestelli M, Yilik L, Ergunesetal K. Recent
surgical experience in chronic constrictive pericarditis.
Tex Heart Inst J 2003;30:27 30.
Inglesis JH, Dec GW. Constrictive pericarditis. Curr
Treat Options Cardiovasc Med 1999;1:63 71.
Rienmuller R, Gurgan M, Erdmann E, Kemkes BM,
Kreutzer E, Weinhold CH. CT and MR evaluation of
pericardial constriction: a new diagnostic and therapeutic concept. J Thorac Imaging 1993;8:108 21.
Havada T, Nakayama K, Kitano T, Sakaguchi M.
Transsternal bilateral thoracotomy for pericardiectomy
after coronary artery bypass grafting. Jpn J Thorac
Cardiovasc Surg 2000;48:480 3.
Uchida T, Bando B, Minatola K, Sasako Y, et al. Pericardiectomy for constrictive pericarditis using the harmonic scalpel. Ann Thorac Surg 2001;72:924 5.
Spodick DH. The normal and diseased pericardium:
current concepts of pericardial physiology, diagnosis
and treatment. J Am Coll Cardiol 1983;1:240 51.
DeValeria PA, Baumgartner WA, Casale AS, et al.
Current indications, risks, and outcome after pericardiectomy. Ann Thorac Surg 1991;52:219 24.
Potwar SA, Arsivala SS, Bhosle KN, Mehta VI. Surgical treatment for chronic constrictive pericarditis. Indian Heart J 1989;4:30 3.
Herrmann G, Gahl K, Simon R, Borst HG, Lichtlen
PR. Pericarditis constriktiva: ergebnisse und probleme
konservativer und operativer Behandlung. Z Kardiol
1983;72:504 13.
Ni Y, von Segeser LK, Turina M. Utility of pericardiectomy for postirradiation constrictive pericarditis?
Ann Thorac Surg 1990;49:445 8.
Levine HD. Myocardial fibrosis in constrictive pericarditis: electrocardiographic and pathologic observations.
Circulation 1973;48:1268 81.
Dines DE, Edwards JE, Burchell HB. Myocardial atrophy in constrictive pericarditis. Mayo Clin Proc 1958;
33:93 9.
Hancock EW. Constrictive pericarditis: clinical clues to
diagnosis. JAMA 1975;232:176 7.
Rienmuller R, Doppman JL, Lissner J, Kemkes BM,
Strauer BE. Constrictive pericardial disease: prognostic
significance of a non-visualized left ventricular wall.
Radiology 1985;156:753 5.
601
[55] Hoit BD. Management of effusive and constrictive

pericardial disease. Circulation 2002;105:2939 84.
[56] Kabbani SS, LeWinter MM. Pericardial diseases. Curr
Treat Options Cardiovasc Med 2002;4:487 95.
[57] Sagrista-Saleda J, Angel J, Sanches A, PermanyerMiralda G, Soler-Soler J. Effusive constrictive pericarditis. N Engl J Med 2004;350:469 75.
[58] Oh KY, Shimizu M, Edwards WD, Tazelaa HD, Danielson GK. Surgical pathology of the parietal pericardium: a study of 344 cases (1993 1999). Cardiovasc
Pathol 2001;10:157 68.
[59] Haley JH, Trajik AJ, Danielson GK, Schaff V, Mulvagh SL, Oh JK. Transient constrictive pericarditis:
causes and natural history. J Am Coll Cardiol 2004;
43:271 5.
[60] Talreja DR, Edwards WD, Danielson GK, Schaff HV,
Tajik AJ, Tazelaar HD, et al. Constrictive pericarditis
in 26 patients with histologically normal pericardial
thickness. Circulation 2003;108:1852.
[61] Giorgi B, Mollet NRA, Dymarkowski S, Rademakers
FE, Bogaert J. Clinically suspected constrictive pericarditis: MR imaging assessment of ventricular septal
motion and configuration in patients and health subjects. Radiology 2003;228:417 24.
[62] Kohli A, Bao KS, Naciappan M, Geelani MA, Balachander J, Bannerjee A, et al. Pericardiectomy for constrictive pericarditis: is prolonged inotropic support a
bad omen? J Indian Med Assoc 2001;99:499 501.
[63] Xiang Y, Du CH, Jiang CQ, He MX, Zhen J, Chen G.
Surgical treatment of constrictive pericarditis : experience in 72 cases. Di Yi Jun Yi Da Xue Xue Bao 2002;
22:279 80.
[64] Bozbuga N, Erentug V, Eren E, Erdigan HB, Kivali K,
Antal A, et al. Pericardiectomy for chronic constrictive tuberculous pericarditis: risk and prediction of survival. Tex Heart Inst J 2003;30:180 5.
[65] Moosdorf R. Indications, results and pitfalls in the
surgery of constrictive pericarditis. Herz 2000;25:
794 802.
[66] Mounsey P. Annular constrictive pericarditis: with an
account of a patient with functional pulmonary, mitral,
and aortic stenosis. Br Heart J 1959;21:325 34.
[67] Ambrosi P, Jau P, Joly P, et al. Intereret de limagerie
par resonance magnetique dans deux cas de constricition annulaire des ventricles. Arch Mal Coeur Vaiss
1991;84:243 7.
Radiol Clin N Am 42 (2004) 603 617
Calcifications of the heart

Ramesh M. Gowda, MDa,b, Lawrence M. Boxt, MDc,d,*
a
The Heart Institute of Beth Israel Medical Center, First Avenue at 16th Street, New York, NY 10003, USA
b
Department of Medicine, Long Island College Hospital, 339 Hicks Street, Brooklyn, NY 11201, USA
c
Department of Radiology, Beth Israel Medical Center, First Avenue at 16th Street, New York, NY 10003, USA
d
Department of Radiology, Albert Einstein College of Medicine of Yeshiva University, 1300 Morris Park Avenue,
Bronx, NY 10461, USA
Calcifications of the heart are in most instances

considered pathologic. They are a marker of increased
risk for cardiovascular morbidity and mortality. Recognizing the shape and location of a calcification is
very useful for diagnosing the underlying disease
causing the calcification. Excess calcium load, chronic inflammation, and malnutrition are associated with
an increased risk for development of cardiac calcifications. Individuals with chronic kidney disease on
maintenance dialysis may experience other metabolic
abnormalities, including hyperphosphatemia, and an
elevated calcium-phosphorus product, increasing the
prevalence of calcification.
Mechanisms postulated in myocardial soft tissue
calcification include carbon dioxide production in
slowly metabolizing tissue (infarcted myocardium);
relative alkalinity; and decreased calcium solubility.
In congenital bicuspid valve, calcification is caused
by constant wear and tear from the abnormal motion
of the valve leaflets. In coronary arteries, calcium
hydroxyapatite deposits early in the formation of
atherosclerotic plaque.
Evaluation of a patient with cardiac calcification is
directed toward detection of calcification within the
heart and pericardium, determining its location and
extent, and its significance. Radiographic examination
of the hearts of autopsied individuals over 40 years of
* Corresponding author. Department of Radiology, Beth

Israel Medical Center, First Avenue at 16th Street, New York,
NY 10003.
E-mail address: lboxt@chpnet.org (L.M. Boxt).
age have shown calcium in over 80% of cases [1].

In more than 50% of these hearts, deposits greater than
3 mm in diameter were found, lesions large enough
for detection by plain film examination.
Detection of coronary and other cardiac calcification is limited by the line pair resolution of the
imaging system and the radiographic technique used
for examination. That is, chest films are obtained
primarily for evaluation of the lungs. High (100
140 kVp) films provide decreased bony detail and
limited detection of parenchymal calcification. Overpenetrated examinations provide better visualization
of calcium, but at the expense of visualizing the
lungs themselves. Fluoroscopic evaluation of the
chest, more commonly performed just before injection of coronary arterial contrast at cardiac catheterization, is more sensitive to the presence of calcium
than plain films. Visual acuity is improved when
searching for the pulsatile motion of calcium in a
beating heart. Furthermore, during fluoroscopy the
patient may be rotated, throwing the center of the
heart, which contains proximal coronary arterial and
aortic valvular calcification, off the spine. Rotating
the chest may bring portions of otherwise nonvisualized pericardial calcification to a heart border-forming location, making its appearance more obvious.
Finally, watching how a solid calcification moves
with the heart and cardiac rotation may confirm its
intracardiac location.
The development of CT significantly added to
sensitivity to the presence of cardiac calcification.
In addition, it readily allows etiologic differentiation
based on location and character of the calcium
deposits. Although limited by long acquisition time,
doi:10.1016/j.rcl.2004.03.010
604
R.M. Gowda, L.M. Boxt / Radiol Clin N Am 42 (2004) 603617
conventional CT became the gold standard for detection of coronary arterial and myocardial, pericardial,
valvular, and intracavitary calcium. The development
of electron-beam CT and the subsequent development
of spiral and now multidetector spiral CT have
decreased image acquisition time to a point where
very small arterial calcium deposits may be detected
and reliably quantitated. Calcium is transparent on a
MR imaging examination, appearing as a signal void
within some other tissue. It is of limited value for
calcium evaluation.
Vascular calcification
Great vessel calcification
Discrete, rimlike calcification of the aorta (Fig. 1)
and aortic arch is usually a sign of degenerative
intimal change, most often resulting from atherosclerosis [2]. The distance between visualized calcification and the outer aortic contour is an estimation of
aortic wall thickness; its measurement may be helpful
in evaluating acute and chronic changes. If care is
taken to consider the effects of supine versus upright
examination, and the effects of tangential visualization of the heart and aorta in oblique or anteroposterior versus posteroanterior examination, then the
distance between intimal calcification and the outer
wall of the aorta should be no greater than 10 mm
[3]. Aortic arch calcification appears curvilinear,
Fig. 1. Anteroposterior ICU radiograph obtained from a

58-year-old man. The entire visualized aorta is characterized
by a thin, intimal calcified layer (arrows). The aorta is
normal in caliber; aortic wall thickness is normal.
often forming a complete ring on frontal views of

the chest. Aortic dissection should be suspected when
the calcification is noted deep to the aortic border
(Fig. 2). Takayasus arteritis, syphilis, sinus of Valsalva aneurysm, and ductus arteriosus calcification
are other etiologies of aortic calcification [4,5]. Location of the calcification plays an important role in
differentiating these various causes (Fig. 3).
Main pulmonary artery calcification may reflect
long-standing, severe pulmonary hypertension [6].
Central pulmonary arterial calcification in patients
with pulmonary hypertension is usually fine and
intimal in nature, producing the appearance of
multiple, round peripherally calcified hilar masses
(Fig. 4). The right ventricular outflow or main pulmonary artery may calcify after infundibulectomy
and outflow patch repair of tetralogy of Fallot
(Fig. 5). Recognition of the pattern of pulmonary
arterial caliber and branching and association with
other typical findings of pulmonary hypertension
(ie, dilatation of the main pulmonary artery [PA] segment) may be helpful for diagnosis.
Coronary artery calcification
There is a strong correlation between coronary
artery calcification and the presence of coronary
atherosclerosis [1]. Arterial calcification is associated
with mural injury, and the evolution of atherosclerotic
plaque formation [7 9]. The incidence of coronary
arterial calcification is similar to the incidence of
atherosclerotic cardiovascular disease in adult individuals. That is, the incidence of coronary arterial
calcification is rare in the second decade of life; it
increases to nearly 100% by the eighth decade. Men
develop coronary arterial calcification nearly a decade earlier than women; the difference in prevalence
of coronary arterial calcification between men and
women is eliminated by the age of 65 to 70 years.
The increased incidence of coronary arterial calcification seen in older individuals parallels the increased
incidence of coronary atherosclerosis found in advancing age. Furthermore, coronary arterial calcification is common in patients with known coronary
artery disease [10,11].
Plain film examination may detect over 40% of
calcified coronary arterial lesions (Fig. 6) detected by
fluoroscopic examination [12]. On plain film examination, coronary arterial calcification presents as curvilinear or tram-track calcification, generally in the
distribution of the proximal coronary tree: to the left
of the geographic center of the heart, adjacent to the
aortic sinuses of Valsalva. Visualization in the PA
film is dependant on radiographic technique, but in an
605
Fig. 2. A 78-year-old woman with acute onset of back pain. (A) Anteroposterior radiograph shows dilatation of the aortic arch
and a curvilinear calcification medial to the lateral border of the arch (arrows). (B) Anteroposterior radiograph obtained 4 years
earlier. The intimal calcification (arrows) is closer to the outer border of the arch.
adequately penetrated film, calcification may be

identified in a triangular region defined by the left
heart border, the spine, and top of the left ventricle
(see Fig. 6A, B) [12]. In lateral view (see Fig. 6C, D;
Fig. 7), calcified arterial segments may be identified
in the distribution over the interventricular septum
(left anterior descending) or anterior atrioventricular
ring (right coronary artery). The calcification usually
has the appearance of a tubular structure, although
involvement of only one wall (resulting in a linear
appearance) may give the appearance of a calcified
myocardial infarction. The association between coronary artery calcification and coronary atherosclerosis
has been observed directly and reported in patients
undergoing coronary angiography (Fig. 8) [13 17].
Coronary angiographic studies have shown that
the presence of coronary calcium is a very sensitive
predictor of occlusive coronary artery disease. In
symptomatic patients referred for catheterization
who underwent fluoroscopy at the time of their
coronary arteriography, 97% of those with calcified
arterial segments on fluoroscopy had at least one
significant ( > 70%) arterial stenosis at catheterization [14].
CT examination is superior to fluoroscopy for
the detection of coronary arterial calcification
[17,18]. Conventional CT (Fig. 9) identifies nearly
twice as many individuals with coronary arterial
calcification than fluoroscopy [19]. In a large multicenter trial, the greater the amount of calcium on
CT scan, the higher the probability of multivessel
obstructive disease found at angiography [20]. Very

fast CT, including electron-beam and spiral and
multidetector spiral CT, have changed the approach
to imaging and quantitating coronary arterial calcification (Fig. 10). Coronary calcium is identified as a
hyperattenuating lesion of at least 130 Hounsfield
units subtending at least three adjacent pixels
Fig. 3. Axial acquisition from a 78-year-old man with

congestive heart failure. Diffuse intimal calcification of the
transverse aortic arch (Ao) is characteristic of atherosclerosis.
606
Fig. 4. A 40-year-old woman with primary pulmonary hypertension. (A) Posteroanterior (PA) radiograph demonstrates dilatation
of the main and hilar pulmonary arteries and right heart. (B) Enlargement of the right hilum reveals curvilinear segmental
pulmonary arterial calcification (arrows).
(each at least 1 mm2). A system for reproducible

coronary calcium scoring [21], based on the area of
detected calcification per coronary cross-section,
multiplied by a factor determined by the maximum
calcium CT density within that cross-section, has
been found to be a valid surrogate for atherosclerotic
plaque burden and as a measure of the severity of
coronary artery disease [4,22 24]. Furthermore, in
prospective studies of asymptomatic individuals
[25,26], coronary calcification detected by electronbeam CT was an independent risk factor for future
coronary events, including myocardial infarction, and
the need for revascularization. Coronary arterial
plaque and coronary arterial calcification, however,
have only a weak correlation with the extent and
distribution of histopathologic stenosis [27]. Although the total atherosclerotic plaque burden was
proportional to the total calcium burden, not all
coronary arterial plaques are calcified, and when
comparing the area of plaque with the area of arterial
calcification [23], calcium area was only about
20% of the total atherosclerotic arterial plaque area.
Perhaps coronary arterial plaque does not calcify
before achieving a certain critical size, or perhaps
calcification present in smaller plaques is not detectible by CT techniques. Calcium scores obtained by
multidetector CT have a high correlation with results
obtained by electron-beam CT [28]. Noninvasive
CT-based evaluation of coronary arteries seems useful in risk stratification of patients with a low to
intermediate pretest likelihood for significant coronary artery disease.
Myocardial calcification
Myocardial calcification is usually classified as
either dystrophic or metastatic [29]. Dystrophic calcification is more common; not associated with elevation of serum calcium or phosphorus levels [30];
Fig. 5. PA radiograph from a 24-year-old woman who is

16 years postrepair of tetralogy of Fallot. Notice the right
aortic arch displacing the trachea toward the left and the
concave main pulmonary artery segment. The fine calcification of the infundibular patch (arrows) defines the top of
the right ventricular outflow.
607
Fig. 6. Asymptomatic 60-year-old man. (A) PA radiograph shows some increased curvature of the left ventricular contour and
mild pulmonary vascular redistribution. A curvilinear calcification is faintly seen (arrows) medial to the mid-left heart border.
(B) Magnified, enhanced view of the mid-left heart border. The large arrow shows the left bronchus crossing the left heart border.
The faintly viewed calcifications are parallel in nature, tram track calcification. (C) Lateral view shows the curvilinear
calcification (arrows) superimposed on the mass of the heart. (D) Magnified, processed view from the lateral examination. The
calcification (arrows) is now seen to be tram track in appearance.
and usually occurs in areas of myocardial necrosis,

hemorrhage, or fibrosis. In older individuals, however, it may be found without focal myocardial abnormality [31]. Myocardial calcification is most
commonly dystrophic, resulting from ischemic heart
disease [32] with myocardial infarction and
scar formation. It is found in 8% of cases of myocardial infarction greater than 6 years old [29]. The
most common site of calcification is in the anterior
wall of the left ventricle. Myocardial calcification
underestimates the size of the underlying myocardial
infarction. Myocardial calcification appears as a thin,
curvilinear calcification, usually found within the
periphery of the infarct, in the distribution of the
interventricular septum and cardiac apex (Fig. 11).

The calcium is distributed away from the aortic root,
and toward the cardiac apex. On posteroanterior
radiograph, myocardial calcification lies to the left
of the midline. Right ventricular calcification is very
rare [29]. Apical left ventricular aneurysms (Fig. 12)
commonly calcify. Isolated extensive papillary muscle calcification is quite a rare finding and seen
following myocardial infarction [33].
Metastatic cardiac calcification is associated with
elevated levels of serum calcium. It is commonly associated with calcium deposition in other tissues,
including the skin, corneas, lungs, stomach, and kidneys [34 37]. Other causes of ventricular calcifica-
608
Fig. 7. Lateral chest examination from a 65-year-old woman

with ischemic heart disease. Notice (arrows) the parallel
curvilinear calcification of the proximal left anterior descending coronary artery.
Fig. 9. Contrast-enhanced CT examination in a patient with

heart failure. Extensive calcification of the anterior descending coronary artery (short arrows) is evident. Pacing wires
(long arrow) pass through the superior vena cava. Intimal
aortic calcification is also seen.
tion include chronic kidney disease and following

cardiac trauma [38]. These calcifications characteristically accumulate in the elastic tissues of the arteries
and the endocardium of the right atrium. Idiopathic
cardiac calcification (ie, calcification without any
underlying etiology) has been reported [39,40] and

is occasionally seen in clinical practice (Fig. 13).
Left atrial calcification is an uncommon manifestation of rheumatic mitral stenosis. When detected, it
is an accurate indicator of rheumatic heart disease.
Fig. 8. Cine frame from a 67-year-old woman with ischemic

heart disease and rheumatic mitral stenosis. The catheter tip
is engaged in the left coronary artery. Notice the calcification
of the anterior descending (short arrows) and distal right
(long arrows) coronary arteries. Thick calcification of the
mitral leaflets (arrowheads) is evident.
Fig. 10. Axial slice from an ECG-gated electron beam CT

examination of the heart of an asymptomatic 40-year-old
man. Faint, distinct calcific plaques (arrows) of the left
anterior descending coronary artery are evident.
609
Fig. 11. Chest examination in a 67-year-old man with a history of previous myocardial infarction. (A) PA radiograph shows
calcification (arrowhead) of a normal size aortic arch. The contour of the left ventricle is rounded and extends toward the left
chest wall. Immediately subjacent to and following the contour is a series of vague calcific densities (arrows). (B) In lateral view,
the fine curvilinear calcification (arrows) appears sharper than in the PA. It follows the bulging interventricular septum.
The calcium is deposited in the left atrial endocardium, more often found posteriorly and superiorly.
Left atrial calcification is usually thin walled and
follows the curvature of the chamber (Fig. 14) [41].
Calcification may also be found within a mural
thrombus in the left atrial appendage. Right atrial
calcification is extremely rare, occurring usually in
the setting of tricuspid rheumatic valvulitis [42]. CT
reveals a laminated structure with calcification and
distinct margins, without invasion of the right atrial
wall [42].
ment for mediastinal malignancy not uncommonly

acquire pericardial calcification characterized by distribution within the port of exposure.
Pericardial calcification may present as a thin focal plaque, or a long, curvilinear layer following the
cardiac contour (Fig. 15). Most pericardial calcifi-
Pericardial calcification
Pericardial calcification results from exposure to
infection, trauma, or hemorrhage, or therapeutic radiation. The most common causes of pericardial
calcification in the past (tuberculosis, histoplasmosis,
and purulent pericarditis) are hardly seen today in the
antibiotic era. Nevertheless, these diseases are still
endemic in other parts of the world, and in an era of
global air travel they may present locally. Traumatic
pericardial disease, including surgical pericardotomy
for intracardiac or coronary artery bypass graft surgery, results in residual blood left within the pericardial space, the nidus for future calcification. Patients
who have undergone previous mantel radiation treat-
Fig. 12. Non contrast-enhanced CT scan from a 67-year-old

man with a prior myocardial infarction. The calcification
follows the periphery of the ventricle, involving the distal
interventricular septum (arrow).
610
Fig. 13. A 34-year-old man with shortness of breath. (A) Overpenetrated display of a PA radiograph demonstrates dense, irregular
calcification projecting over the dilated left ventricle. (B) On lateral examination the dense calcification is projected over the
ventricular mass. Notice the posterior displacement of the left ventricular wall (arrows) indicating dilatation. (C) Oblique axial
double inversion recovery MR acquisition. Although the left ventricle (LV) is hypertrophied, there are numerous irregular signal
voids throughout the myocardium (arrows) representing the endocardial and myocardial calcifications.
cation is only about 1 to 2 mm in thickness, but longstanding disease may be associated with 1- to 2-cm
thick lesions. Pericardial calcification is most commonly found within the atrioventricular grooves (dependent portions of the pericardial space) and in
the lower and diaphragmatic portions of the pericardium. Pericardial calcification is usually found on
both the right and left sides of the heart. Although
it may present as a local plaque, it more commonly
is seen as an extensive process. Differentiation between myocardial and pericardial calcification is
based on the distribution and character of the calcification. Pericardial calcification tends to be diffuse,
globally involving the pericardial space, and surrounding the heart (Fig. 16). Myocardial calcification
localizes to the left side of the heart; the myocardium
resides to the left. Differentiating a solitary pericardial from myocardial calcification based on the peripheral distribution of the pericardium may be
difficult on CT examination, and nearly impossible
on a plain film. Pericardial calcification tends to be
clunky and ugly in character (Fig. 17), whereas
611
[43,44]. Calcification is an important sign of pericardial constriction, but is not pathognomonic. Pericardial calcification may be present in the absence of any
physiologic insult to the heart.
Valvular calcification
Fig. 14. Axial contrast-enhanced spiral CT acquisition from

a 54-year-old woman with rheumatic mitral stenosis. The
markedly dilated left atrium (LA) extends from right heart
border to left. Thick, peripheral calcification indicates
rheumatic heart disease. Note that the right atrial appendage
(RAA) and main pulmonary artery (PA) are both greater in
caliber than the ascending aorta (Ao).
myocardial calcification tends to be fine and curvilinear (see Fig. 12).

Pericardial calcification indicates the diagnosis of
calcific pericarditis. The association between pericardial calcification and constrictive pericarditis is not
constant. Between 30% and 70% of patients with
constrictive pericarditis have pericardial calcification
Valvular calcification usually indicates the presence of valvular sclerosis or hemodynamically significant stenosis [45]. It is commonly associated with
rheumatic fever, congenital malformation, old endocarditis, and atherosclerosis. Mitral valvular calcification is overwhelmingly associated with preceding
rheumatic fever. Mitral valve calcification can take
two forms. Mitral annular calcification is rarely seen
before the sixth decade and is considerably more
common in women (Fig. 18). Annular calcification
appears as dense ringlike clumps, varying from 2 to
4 cm in diameter. The ring takes a particular orientation, defining the posterior atrioventricular ring.
Annular calcification is commonly associated with
normal mitral valve function (Fig. 19). When present,
however, the valvular dysfunction is more often
mitral insufficiency than stenosis. Mitral annular
calcification is common in end-stage renal disease
and may develop and progress over a short period of
time. Mobile components associated with mitral annulus calcification detected by echocardiography may
directly cause cerebrovascular accidents [46]. Mitral
Fig. 15. A 96-year-old woman who feels a little tired. (A) PA radiograph shows the long, irregular peripheral calcification
(arrows) along the right heart border, extending to beneath the heart. (B) In lateral examination, the calcification (arrows) follows
the anterior aspect of the heart.
612
Fig. 16. Planar reconstructions obtained from a 70-year-old woman with shortness of breath. (A) Coronal reconstruction obtained
immediately posterior to the sternum (long arrow). The calcified parietal (arrow 1) and visceral (arrow 2) pericardial layers are
separated by serous fluid. Note the right pleural effusion (eff), and immediately inferior to the right diaphragm, ascites (asc).
(B) Coronal reconstruction 3 cm behind Fig. 16A. The calcified parietal pericardium extends up to the ascending aorta (Ao) on
the right, and over the top of the main pulmonary artery (PA) on the left.
leaflet calcification may be delicate and difficult to

detect (Fig. 20). Leaflet involvement is characteristic
of rheumatic mitral stenosis.
Aortic valve calcification is associated with aortic
stenosis. Purely regurgitant valves and nonstenotic
congenitally bicuspid valves tend not to calcify [47].

Aortic valve calcification detected before the fourth
decade is usually in a congenitally bicuspid valve;
calcification detected after the sixth decade is associated with acquired valve degeneration.
Fig. 17. Axial electron beam CT acquisition through the

diaphragmatic surface of the heart in a 67-year-old woman
with pericardial constriction. The inferior aspect of the left
ventricular (LV) cavity and upper inferior vena cava (IVC)
are labeled. The dense, irregular pericardial calcification
(arrows) covers most of the inferior cardiac surface.
Fig. 18. Lateral chest radiograph from a healthy 83-year-old

woman. The left atrium and ventricle are normal. Heavy, Cshaped calcification (arrows) in the posterior atrioventricular
ring is typical of degenerative mitral annular calcification.
Fig. 19. Oblique sagittal planar reconstruction from a

contrast-enhanced spiral multidetector CT examination of a
76-year-old man with bronchogenic carcinoma. The right
pleural effusion (eff) is labeled. The markedly calcified
mitral annulus separates the normal-size left atrium (LA)
from a hypertrophied left ventricle (LV). Note the calcified
descending aorta (arrow).
613
third of the lateral cardiac silhouette. In congenital

bicuspid valve stenosis the calcification usually
appears as a thick, irregular semilunar ring with a
central knob (Fig. 22), resulting from calcification of
the valve raphe. Occasionally, fusion of two of the
three leaflets of a normal trileaflet aortic valve from
calcification may appear like a bicuspid valve. Degenerative calcific aortic stenosis (acquired aortic
stenosis) is most often found in patients with tricuspid
aortic valves [48]. Calcification involves the annulus
and leaflets and may be heavy.
Use of spiral CT allows differentiation between
leaflet and aortic annular calcification, potentially
providing a means of differentiating clinically significant stenosis from degenerative annular calcification,
a process associated with aging (Fig. 23).
Tricuspid and pulmonary valve calcification is
very unusual. The tricuspid valve lesion may be
associated with long-standing rheumatic heart disease. Pulmonary valve calcification may be found in
older adult patients, associated with valvular pulmonary stenosis.
Intracavitary calcification
The aortic valve resides in nearly the geographic
center of the heart. It is often projected over the spine,
limiting the value of posteroanterior radiography for
its detection (Fig. 21). In lateral view, aortic valve calcification is thick, and often found within the middle
Tumor calcification
The most common tumor of the heart, the left
atrial myxoma, calcifies in about 10% of cases
Fig. 20. Frames of a cineangiogram in cranialized left anterior oblique projection obtained from the same patient in Fig. 8.
(A) End diastolic frame shows separation of the thickened, calcified anterior (A) and posterior (P) mitral leaflets. (B) The two
leaflets coapt during ventricular systole. Note the limited excursion of the leaflets, reflecting the limited mitral orificial area in
mitral stenosis.
614
Fig. 21. A 54-year-old man with degenerative calcific aortic stenosis. (A) PA radiograph shows increased curvature of the lower
left heart border, and dilatation of the ascending aorta (arrows). The aortic arch (Ao) is not dilated. (B) Lateral view shows the
dense, thick calcification in the center of the heart (short arrows). The retrosternal clear space is filled from behind (arrows
a,b,c) by the dilated ascending aorta. Also note the clear inferior retrocardiac space just above the gastric air bubble (Bu). The left
ventricle is hypertrophied, but not dilated.
[49,50]. The distribution of the calcification is high

and posterior in the heart, reflecting its containment
within the left atrium. The pattern of calcification is
speckled and central in location, probably reflecting
a necrotic core (Fig. 24). Motion of the calcified
mass prolapsing through the mitral orifice may be

appreciated by fluoroscopic examination. Other cardiac tumors may calcify, including fibromas and
rhabdomyomas. These masses are more commonly
found in infancy and childhood. Metastatic cardiac
Fig. 22. Electron beam CT acquisition from a 30-year-old

man with a stenotic congenital bicuspid aortic valve. Axial
image through the calcified (arrow) central portion of the
valve. Small, bilateral pleural effusions (eff) are evident.
Fig. 23. Acquisition from a 64-year-old man with aortic

stenosis. The aortic valve annulus and leaflets (arrows) is
calcified. Notice the aortic and posterior pleural calcification.
615
Fig. 24. A 70-year-old woman with intermittent shortness of breath. (A) PA radiograph shows flattening of the left atrial
appendage segment of the left heart border (arrow), indicating left atrial enlargement. (B) In lateral view, the irregularly calcified,
3-cm left atrial myxoma is seen (arrows) abutting the posterior atrioventricular ring, and mitral orifice.
tumors are more common than primary cardiac malignancies, but these tumors do not calcify sufficiently for plain film diagnosis.
Summary
Cardiac calcification may be a reflection of degenerative processes associated with aging, and not a
reflection of a pathologic process that affects cardiac
function. This is probably true in cases of mitral
annular and isolated aortic annular calcification. Pericardial calcification indicates a previous insult; in the
proper clinical circumstances, pericardial calcification
indicates pericardial constriction, a clinically important condition to exclude. Calcification of the coronary arteries reflects the presence and progression of
atherosclerosis. Use of very fast CT allows quantitation of coronary calcium, a method of screening
patients at risk for coronary heart disease. Aortic
leaflet calcification is associated with a valvular
gradient. Myocardial calcification reflects the presence of a scar or ventricular aneurysm.
References
[1] Blankenhorn DH, Stern D. Calcification of the coronary arteries. AJR Am J Roentgenol 1959;81:772 7.
[2] Edwards JE. Manifestations of acquired and congenital
diseases of the aorta. Curr Probl Cardiol 1979;3:1 62.
[3] Crawford ES. The diagnosis and management of aortic

dissection. JAMA 1990;264:2537 41.
[4] Rubin GD. Helical CT angiography of the thoracic
aorta. J Thorac Imaging 1997;12:128 49.
[5] Sharma S, Rajani M, Talwar KK. Angiographic morphology in nonspecific aortoarteritis (Takayasus arteritis): a study of 126 patients from north India.
Cardiovasc Intervent Radiol 1992;15:160 5.
[6] Roberts HC, Kauczor HU, Schweden F, et al. Spiral
CT of pulmonary hypertension and chronic thromboembolism. J Thorac Imaging 1997;12:118 27.
[7] Hirota S, Imakita M, Kohri K, et al. Expression of
osteopontin messenger RNA by macrophages in atherosclerotic plaques: a possible association with calcification. Am J Pathol 1993;143:1003 8.
[8] Ideta T, Shirasawa T, Esaki Y, et al. Osteopontin
mRNA is expressed by smooth muscle-derived foam
cells in human atherosclerotic lesions of the aorta.
J Clin Invest 1993;92:2814 20.
[9] Shanahan CM, Nary NR, Metcalfe JC, Weissberg PL.
High expression of genes for calcification-regulating
proteins in human atherosclerotic plaque. J Clin Invest
1994;93:2393 402.
[10] Jorgens J, Boardman WJ, Damberg SW, et al. The significance of coronary calcification. AJR Am J Roentgenol 1965;95:667 72.
[11] McCarthy JH, Palmer FJ. Incidence and significance
of coronary artery calcification. Br Heart J 1974;36:
499 506.
[12] Souza AS, Pream PR, Elliott LP. Chest film detection
of coronary arterial calcification: the value of the CAC
triangle. Radiology 1978;129:7 10.
[13] Aldrich RF, Brensike JF, Battaglini JW, et al. Coronary
calcification in the detection of coronary artery disease
and comparison with electrocardiographic exercise
testing. Circulation 1979;59:113 24.
616
[14] Bartel AG, Chen JTT, Peter RH, et al. The significance
of coronary artery calcification detected by fluoroscopy: a report of 360 patients. Circulation 1974;49:
1247 53.
[15] Kelley MJ, Huang EK, Langou RA. Correlations of
fluoroscopically detected coronary artery calcification
with exercise stress testing in asymptomatic men. Radiology 1978;129:1 6.
[16] Margolis JR, Chen JTT, Kong Y, et al. The diagnostic
and prognostic significance of coronary artery calcification. Radiology 1980;137:609 16.
[17] Schultz KW, Thorsen MK, Gurney JW, et al. Comparison of fluoroscopy, angiography and CT in coronary
arterial calcification. Appl Radiol 1989;6:38 43.
[18] Stanford W, Rooholamini M, Rumberger J, et al. Evaluation of coronary bypass graft patency by ultrafast
computed tomography. J Thorac Imaging 1988;3:
52 5.
[19] Moore EH, Greenberg RW, Merrick SH, et al. Coronary artery calcifications: significance of incidental
detection on CT scans. Radiology 1989;172:711 6.
[20] Budoff MJ, Georgiou D, Brody A, et al. Ultrafast computed tomography for the detection of coronary artery
disease: a multicenter study. Circulation 1996;93:
898 904.
[21] Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary calcium using ultrafast computed
tomography. J Am Coll Cardiol 1990;15:827 32.
[22] Blumgart D, Schmermund A, Goerge G, et al. Comparison of electron beam computed tomography with
intracoronary ultrasound and coronary angiography for
detection of coronary atherosclerosis. J Am Coll Cardiol 1997;30:57 64.
[23] Rumberger JA, Simons DB, Fitzpatrick LA, et al.
Coronary artery calcium areas by electron beam computed tomography and coronary atherosclerotic plaque
area: a histologic correlative study. Circulation 1995;
92:2157 62.
[24] Schmermund A, Denktas AE, Rumberger JA, et al.
Independent and incremental value of coronary artery
calcium for predicting the extent of angiographic coronary artery disease: comparison with cardiac risk factors and radionuclide perfusion imaging. J Am Coll
Cardiol 1999;34:777 86.
[25] Arad Y, Spadaro LA, Goodman K, et al. Predictive
value of electron beam CT of the coronary arteries:
19-month follow-up of 1173 asymptomatic subjects.
[26] Detrano RC, Wong ND, Tang W, et al. Prognostic
significance of cardiac cinefluoroscopy for coronary
calcific deposits in asymptomatic high risk subjects.
J Am Coll Cardiol 1994;24:354 8.
[27] Sangiorgi G, Srivatsa SS, Rumberger JA, et al. Arterial
calcification and not lumen stenosis is highly correlated with atherosclerotic plaque burden in humans;
a histologic study of 723 coronary artery segments
using nondecalcifying methodology. J Am Coll Cardiol 1998;31:126 33.
[28] Stanford W, Thompson BH, Burns TL, et al. Coronary
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
artery calcium quantitation at multi-detector row helical CT versus electron-beam CT. Radiology 2004;230:
397 402.
Freundlich IM, Lind TA. Calcification of the heart and
great vessels. CRC Crit Rev Clin Radiol Nucl Med
1975;6:171 216.
Bloor CM. Cardiac pathology. Philadelphia: JB Lippincott; 1978.
Waller BF, Morgan R. The very elderly heart. In: Waller BF, editor. Contemporary issues in cardiovascular
pathology. Philadelphia: FA Davis; 1988. p. 361.
Lasser A. Calcification of the myocardium. Hum Pathol
1983;14:824.
Schwender FT. Papillary muscle calcification after
inferoposterior myocardial infarction. Heart 2001;
86:E8.
Fierer JA, Wagner BM, Strebel RF. Metastatic calcification of the myocardium. Am J Cardiol 1970;
26:423 6.
Kuzela DC, Huffer WE, Conger JD, et al. Soft tissue
calcification in chronic dialysis patients. Am J Pathol
1977;86:403 24.
Mulligan RM. Metastatic calcification. Arch Pathol
1947;43:177 230.
Parfitt AM. Soft tissue calcification in uremia. Arch
Intern Med 1969;24:544 56.
Jing J, Kawashima A, Sickler A, et al. Metastatic cardiac calcification in a patient with chronic renal failure
undergoing hemodialysis: radiographic and CT findings. AJR Am J Roentgenol 1998;170:903 5.
Duke M. Massive calcification of myocardium of
unknown origin. Arch Pathol 1957;64:34 8.
Patterson D, Gibson D, Gomes R, et al. Idiopathic calcified myocardial mass. Thorax 1974;29:
589 94.
Leonard JJ, Katz S, Nelson D. Calcification of the left
atrium: its anatomic location, diagnostic significance
and roentgenologic demonstration. N Engl J Med
1957;256:629 34.
Aoyagi S, Nishimi M, Hiratsuka R, Takaseya T, Teshima H. Right atrial thrombus associated with combined valvular disease: case report. J Heart Valve Dis
2001;10:542 4.
Cornell SH, Rossi NP. Roentgenographic findings in
constrictive pericarditis: an analysis of 21 cases. AJR
Rigler LG, Wangensteen OH, Friedell HL. Roentgen
kymography in constrictive pericarditis. AJR Am J
Roentgenol 1941;46:765 72.
Rodan BA, Chen JTT, Halber MD, et al. Chest roentgenographic evaluation of the severity of aortic stenosis. Invest Radiol 1982;17:453 7.
Willens HJ, Ferreira AC, Gallagher AJ, Morytko JA.
Mobile components associated with rapidly developing
mitral annulus calcification in patients with chronic
renal failure: review of mobile elements associated
with mitral annulus calcification. Echocardiography
2003;20:363 7.
Roberts WC. The congenitally bicuspid aortic valve:

a study of 85 autopsy cases. Am J Cardiol 1970;26:
72 83.
[48] Roberts WC, Perloff JK, Constantino T. Severe valvular
aortic stenosis in patients over 65 years of age: a clinicopathologic study. Am J Cardiol 1971;27: 497 500.
617
[49] Reynen K. Cardiac myxomas. N Engl J Med 1995;33:

1610 7.
[50] White CS. MR evaluation of the pericardium and cardiac malignancies. Magn Reson Imaging Clin N Am
1996;4:237 51.
Radiol Clin N Am 42 (2004) 619 634
Cardiac imaging using nuclear medicine and postitron

emission tomography
Andrew Crean, MRCP, FRCRa, David Dutka, FRCPb,
Richard Coulden, FRCP, FRCRa,*
a
Department of Radiology, Papworth Hospital, Cambridge, CB3 8RE UK

b
Cardiology, Addenbrookes Hospital, Cambridge, UK
An article of this type cannot hope to address all

aspects of cardiac nuclear medicine for the novice;
therefore, a basic knowledge of principles and technique has been assumed. The article concentrates
on specific issues that are of current interest in
mainstream nuclear cardiology. These include developments in myocardial perfusion technique, the
potential diagnostic benefits of ECG-gating and attenuation correction, nuclear imaging in the diagnosis of
hibernating myocardium, and the cost-effectiveness of
perfusion imaging in patients with suspected angina.
Myocardial perfusion technique

Myocardial perfusion imaging has a well-established track record but in many countries has failed to
gain a central role in cardiology decision-making.
The reasons for this vary around the world but there
are some common themes. Two that are particularly
important are limited access to nuclear perfusion
imaging, and lack of confidence in the report.
Nuclear perfusion imaging must be freely available and performed in an appropriate time scale. Any
delay is likely to result in the referring clinician
opting straight for coronary angiography. The report
Dr. Crean receives grant support from the Royal

College of Radiologists UK. Some of the images shown
were produced during ongoing research funded by this body.
E-mail address: rc@prdg.demon.co.uk (R. Coulden).
must not only be reliable but also presented in a

systematic format that answers the clinical question.
As demand for perfusion imaging grows, how does
one optimize efficiency and increase patient throughput? Many departments are already working to capacity. How does one standardize reports and reduce
interreporter variability?
Exercise or pharmacologic stress?
Physician-supervised exercise stress is still widely
used for nuclear perfusion imaging. Indeed, most
national guidelines promote treadmill stress as the
preferred technique. There is no doubt that exercise
stress is more physiologic than pharmacologic stress
but there is no evidence that this equates with a
higher diagnostic accuracy [1]. For patients who are
unable or unwilling to exercise, there is a significant
risk of obtaining false-negative results or having to
repeat the examination with pharmacologic stress.
When an examination has to be repeated this disrupts
department schedules, and increases costs and radiation burden to the patient.
There is a strong argument for moving toward
pharmacologic stress in all patients. Pharmacologic
stress is fast, safe, and most importantly reproducible.
Variations in the level of exercise stress achieved,
either through patient cooperation or the expertise of
the supervising doctor, are eliminated. The safety
profile of the pharmacologic stress agents is well
established and the available data suggest that there is
no excess of adverse incidents when the stress agent
is administered by a technician. In an audit of local
practice, medical intervention was needed during
doi:10.1016/j.rcl.2004.03.002
620
A. Crean et al / Radiol Clin N Am 42 (2004) 619634
Table 1
Frequency of side effects seen with pharmacologic stress in the authors institution over a 4-year period
No symptoms %
Minor or moderate %
Early termination %
Medical intervention %
Adenosine (850)
Dobutamine (261)
Dipyridamole (86)
5.5
85.5
7.9
0.5
9.2 (24)
64 (167)
23.8 (62)
4.5 (12)
22 (19)
48 (41)
NA
30 (26)
(47)
(727)
(67)
(4)
Numbers in brackets refer to the number of patients in each group. Dipyridamole was abandoned after the first 2 years because of
a high level of side effects.
adenosine stress in 0.5% of cases and during dobutamine stress in 5% (Table 1) [2]. In none of these did
intervention amount to more than plasma expansion
using a colloid infusion or use of a nitrate spray. A
duty doctor must be available in the case of an
adverse reaction or prolonged symptoms, but does
not need to be physically present. This simplifies
bookings when supervising junior medical staff are in
short supply, improving efficiency for the department
and the duty doctor [3]. When using vasodilating
stressors, such as adenosine or dipyridamole, there is
no need to stop rate-limiting antianginal medication.
b-Blockers are only stopped for dobutamine stress
tests (24 hours before examination).
One- or two-day protocol?
Thallium 201 (201Tl) imaging is by its very nature
a 1-day protocol. As a potassium analogue, it is taken
up rapidly by myocytes with a high first-pass extraction fraction (85%). Initial myocardial uptake is
directly proportional to flow up to approximately
2.5 times resting blood flow. Because 201Tl is not
bound to myocytes, redistribution starts immediately.
Rapid redistribution means stress imaging must be
performed as soon after injection as possible, before
significant redistribution occurs. Rest images are
acquired later, usually at 2 to 4 hours. This ties the
patient stressing process closely to the gamma
camera room, increasing camera time and potentially
reducing patient throughput. When one adds the
disadvantages of long half-life, high patient radiation
dose (18 mSv), and a gamma emission energy that
is relatively poor for imaging, it is clear why 201Tl
is used less and less. Sixty percent of departments
in the United Kingdom now use technetium
99m (99mTc) based agents.
By contrast, 99mTc is better suited to perfusion imaging. 99mTc has a relatively short half-life
(6 hours); has a gamma emission energy of 140 KeV,
which is ideally suited to the physics of the gamma
camera; better radiation dosimetry (allowing a higher
administered dose for less absorbed dose [6 mSv]);

and higher count statistics (giving better image quality). Two 99mTc agents are currently available: sestamibi and tetrofosmin. Sestamibi has lower first-pass
extraction than 201Tl but uptake remains proportional
to flow up to two times resting levels. At flow rates
above this, uptake reaches a plateau (Fig. 1). Unlike
201
Tl, sestamibi is bound irreversibly to myocyte
mitochondria. This has the important consequence
that tracer distribution, when imaged hours later,
mirrors perfusion at the time of injection. Imaging
may be delayed by 1 to 2 hours, separating the stress
process from the gamma camera room. This delay
also allows time for tracer to be cleared from the liver
(a major target organ for sestamibi) resulting in
improved cardiac image quality [4]. Tetrofosmin is
the newest of the 99mTc agents with similar myocardial uptake, retention, and clearance to sestamibi.
Faster clearance from lung and liver, however, gives
some benefit over sestamibi.
Because sestamibi is bound irreversibly to myocardium, stress and rest imaging require separate
injections of tracer. If the second image acquisition
Fig. 1. Graph showing relationship between myocardial

blood flow and uptake of a variety of radionuclide tracers.
Comparison is made with microspheres that show a linear
relationship. Thallium provides a linear response to increasing flow to 250% of resting flow. This is a greater range
than any of the technetium 99m tracers.
is not to be contaminated with radiation from the first,

there needs to an interval between the two injections
of at least four half-lives (24 hours) (ie, a 2-day
protocol). This is inconvenient for the patient who
has to attend the department on separate days but is
relatively radiation dose efficient. A 1-day protocol is
achieved through using a low dose injection for stress
imaging (250 MBq) with a much larger injection for
the rest examination (750 MBq) several hours later.
Radiation from the larger second dose saturates the
effects of the decaying stress injection thereby giving
satisfactory rest images. The stress examination needs
to be performed first. If the rest examination is first,
there is a risk that perfusion defects produced by the
later stress injection may be masked by residual
background radiation from the rest study. Although
convenient for the patient, the total administered
activity for a 1-day protocol (1000 MBq) is higher
than that for the 2-day (800 MBq given as two
400 MBq injections). The low level of injected activity for the stress acquisition using the 1-day
protocol is inadequate for ECG-gated single-photon
emission CT (SPECT).
The accuracy of sestamibi-based perfusion imaging has been shown to be at least as good as 201Tl in
numerous studies. Given the benefits detailed, there is
little doubt its use should be the norm. Is a 2-day
sestamibi protocol, however, really necessary? Is a
rest examination essential? A number of authors have
proposed a schema in which patients undergo the
stress examination on 1 day, but do not proceed to the
rest study on the second day if it is normal. Gibson
et al [5] showed that, in patients with low to medium
pretest probability of coronary artery disease (CAD),
a normal stress perfusion study predicts a subsequent
cardiac event rate of 0.6%. This compares well with
event rates seen in large studies of similar patient
groups who have undergone both stress and rest
imaging. In another smaller study, stress-only SPECT
was found to be capable of detecting and localizing
CAD as accurately as the stress-rest protocol [6].
Depending on the proportion of normal examinations
in the population being studied, this has major implications for both cost and patient throughput. Even
in the United Kingdom, where the prevalence of disease being investigated is high, this is likely to affect
30% of cases, freeing gamma camera time for more
patient examinations.
Reporting of results
Unfortunately, the quality of nuclear perfusion
reports is highly variable. Not all centers meet the
levels of accuracy and reproducibility that are
621
expected from the literature. Nuclear perfusion reports need to be systematic and should include the
following main points:
1. Type of stress and if completed
2. Physiologic response to stress
3. Patient parameters if likely to be a cause of
attenuation artifact
4. Location, severity, and reversibility of reductions in tracer uptake
5. Additional findings (eg, transient ventricular
dilatation)
It is important to mention the method of stress
used and whether the patient experienced symptoms
or had ECG changes during or after stress. Given
the potential for antagonism of heart rate rise by
b-blockers and adenosine by caffeine, it is essential
to include details of change in pulse and blood
pressure. Any patient who experiences little or no
physiologic change may have had inadequate stress,
so the report should be qualified accordingly (risk
of a false-negative result). One must be particularly
cautious when a patient with a high pretest probability of CAD has no symptoms, no change in
physiologic parameters, and a normal scan. In this
case it may be necessary to repeat the examination
with exercise or dobutamine stress as appropriate.
Patient height, weight, and body habitus should be
considered because these have a bearing on attenuation artifact.
Tomographic images are best reviewed in standard cardiac planes (Fig. 2). Some authors prefer a
monochromatic gray scale, whereas others have argued that a continuous color scale provides better
interobserver agreement [7]. More important is consistency of approach. Reporters should become familiar with one display system and adhere to that
system. The location, severity, and reversibility of all
perfusion abnormalities should be described.
Location is best dealt with using a standard
segmental model and the 17-segment model, recently
adopted by consensus among a number of imaging
bodies, is ideal (Fig. 3) [8]. Description of location
and its relationship to the segmental model is helped
by using bulls-eye plots of rest and stress data
(Fig. 4). Tracer uptake in the stress portion of the
examination should be described in semiquantitative
terms (ie, normal, mildly reduced, moderately reduced, or severely reduced). When scaling the display
to show the most intense region of myocardial uptake
as 100%, it must be remembered that normal variations in perfusion can reduce tracer uptake by 30%.
Even greater reductions can be seen in regions of
622
Fig. 2. (A) Standard cardiac tomographic planes derived from SPECT data showing short axis in rows from apex to base with
stress images above matched to rest images below. Horizontal and vertical long axis slices are also matched stress with rest.
(B) Short axis rings of left ventricular activity can be nested one inside the other with basal ring at the periphery and apex at
the centre to give a bulls-eye plot. Bulls-eye plots showing rest (C) and stress (D) from a sestamibi SPECT data set in
a normal patient.
normal myocardial perfusion when attenuation artifact is present. The reversibility of defects compared
with rest images needs to be described semiquantitatively (ie, fully reversible, partly reversible, or irreversible) (Fig. 5). In some instances, a pattern of
reverse distribution may occur with areas of normal
uptake during stress showing patchy decrease in
tracer activity at rest. Although it has been suggested
that this appearance may be caused by partial thick-
ness infarction in the presence of a patent subtending

artery, its true significance is unknown [9,10].
Artifacts should be sought and their cause described. Most artifacts are caused by attenuation: the
anterior wall in women because of breast attenuation
and the inferior wall in men from the diaphragm
(Fig. 6). It may be possible to see the attenuating
structure when viewing a rotating cine of the projection images, although the same conclusion may be
Fig. 3. Seventeen-segmental model for description of left

ventricular regions as defined by consensus. (From Cerqueira
M, Weissman N, Dilsizian V. Standardized myocardial segmentation and nomenclature for tomographic imaging of the
heart. Circulation 2002;105: 539 42; with permission.)
drawn more simply by reviewing details of the

patients body habitus and bra size. Artifacts caused
by patient motion or extracardiac activity (eg, liver)
also are picked up from the projection data. In
general, attenuation artifacts appear as fixed perfusion defects (their cause being the same on both rest
623
and stress data sets). Occasionally, large breasts may

lie in different positions on rest and stress imaging
and breast strapping or prone imaging may be necessary. When a fixed defect is suspected as being
caused by attenuation, wall motion on ECG-gated
images should be normal. Impaired wall motion
suggests infarction and that the defect is genuine.
Reconstruction artifacts are less common, although
the phenomenon of normal apical thinning is generally well recognized [11].
Having assessed myocardial tracer uptake, the
reporter should look for transient ischemic dilation
of the left ventricle and lung uptake. Both of these
findings are far more likely to occur when using
thallium but have also been described with 99mTc
tracers. Transient ischemic dilation is a poor prognostic indicator associated with a high risk of perioperative and long-term cardiac events (Fig. 7) [12].
The mechanism for this when 201Tl is used for
imaging remains uncertain. Postischemic stunning
can certainly produce ventricular dilatation, although
overestimation of ventricular cavity size through
subendocardial ischemia and poor border definition
are also likely to be important. When using 99mTc
agents, transient ischemic dilation is of particular
significance because the stress study is usually acquired 60 minutes or more following stress. This
compares with 5 to 10 minutes for 201Tl. Ventricular
dilatation indicates severe ischemia with stunning.
Like transient ischemic dilation, the presence of lung
Fig. 4. Bulls-eye plots showing rest (A) and stress (B) sestamibi SPECT in a patient with large, moderately severe, but almost
fully reversible defect in the mid and distal anterior wall and apex.
624
Fig. 5. Bulls-eye plots showing rest (A) and stress (B) sestamibi SPECT in a patient with large, moderately severe, fixed defect in
the mid and distal anterior wall and apex.
uptake is a strong indicator of future cardiac events. It

is not, however, specific to ischemia. Elevated left
atrial pressure from other causes, such as cardiomyopathy or severe valve disease, also give rise to
this phenomenon.
ECG-gated single-photon emission CT imaging

One of several advantages of using a 99mTc agent
compared with 201Tl is that count statistics are high
enough to allow an ECG-gated acquisition. Instead of
collecting SPECT data continuously, independent
of the ECG, data collection is divided into a number
of equal time periods or phases throughout the cardiac cycle (Fig. 8). Data for each phase are held
separately and can be used either alone or together to
reconstruct standard tomographic slices. Used together, images are identical to those obtained from a
nongated acquisition. Constructed separately, the tomographic slices depict myocardial uptake at multiple
time points through the cardiac cycle. When viewed as
a cine loop, each tomographic slice can be used to
assess wall motion and ventricular function. For a
standard gamma camera, ECG-gating adds approximately 30% to the acquisition time. This equates to an
increase from 10 to 13.5 minutes for the average dualhead system. The time penalty may be impractical on
older single-head units, where this represents an increase from 20 to 30 minutes, but the benefits far
outweigh the time cost on a modern system.
Those institutions that use ECG-gating routinely

normally acquire eight phases per cardiac cycle. There
is evidence that sampling 16 phases per cycle results
in a more accurate calculation of ejection fraction
[13], but the distinction is rarely important when
good, moderate, or poor is more than sufficient for making a clinical decision in most cases.
Like all other aspects of nuclear perfusion imaging, attention to detail is essential. Technicians need
to be aware of the importance of a good quality ECG
and both the beat-length histogram and the left
ventricular time-volume curve should be examined
to assess gating accuracy. Many software packages
use automatic edge detection for wall motion analysis
but care should be taken to ensure that computerdefined edges are appropriate. Errors lead to overestimation or underestimation of ventricular volumes
and ejection fraction. Comparison of ventricular
volumes and ejection fraction derived from ECGgated SPECT, radionuclide ventriculography [13],
and echocardiography show good correlation. Absolute volumes, however, tend to be underestimated
because of the poor spatial resolution of gamma
camera systems.
Does gated single-photon emission CT add value?
ECG-gated 99mTc SPECT is unequivocally better
than 201T SPECT for identifying CAD and holds true
whether using sestamibi [14] or tetrofosmin [15].
This is explained by the added value of ECG-gated
625
Fig. 6. Bulls-eye plots showing rest (A) and stress (B) sestamibi SPECT in a patient with a mild fixed defect in the mid and distal
anterior wall. This is a typical site for breast attenuation artifact and showed normal wall motion on ECG-gated SPECT. (C) Left
coronary angiogram confirms normal left anterior descending.
images in displaying regional wall motion. When

faced with fixed perfusion defects that look real but
are in a typical location for attenuation artifact,
preserved wall motion and systolic wall thickening
favor artifact over infarct. As a result, routine use of
ECG-gated SPECT enhances diagnostic confidence.
In a study of 285 consecutive patients, ECG-gating
decreased the number of equivocal reports (borderline normal or abnormal) from 31% to 10%. Furthermore, patients with a low pretest probability of
disease were more likely to be classified as normal
(74% increasing to 93%) once ECG-gated images
were reviewed [16]. Improved accuracy and diagnostic confidence has had a major impact on the position
of nuclear perfusion imaging in the diagnostic
chain. In patients with a low or intermediate probability of CAD, ECG-gated SPECT now becomes a
cost-effective gatekeeper for coronary angiography.
Quantitative variables derived from ECG-gated
images have also been shown to contain powerful
and additional prognostic information. Sharir et al
[17] have shown that poststress ejection fraction is a
risk factor in predicting cardiac death and that this is
independent of the extent of the perfusion defect. In
626
Fig. 7. Bulls-eye plots showing rest (A) and stress (B) sestamibi SPECT in a patient with a large severe reversible defect in the
mid and distal anterior apex and anterior septum wall. This defect is associated with transient ischemic dilatation of the left
ventricle as shown in all three planes at rest (C) and (D) stress (1 hour after stress injection). Figs. 7C and D are ECG-gated
SPECT images in systole. Arrows show focal area of hypokinesia with stress.
Fig. 8. Diagram showing process for data acquisition for

ECG-gated SPECT. As gamma camera head rotates around
the patient, multiple sets of data are acquired at different
time points through the cardiac cycle. All data sets for the
same time point are used to construct a set of tomographic
slices, with the process repeated for each time point. Each
tomographic slice is represented by multiple images
throughout the cardiac cycle, which can be viewed as a
cine loop to assess wall motion and thickening.
another study of 2500 patients, the authors found that

poststress ejection fraction (ECG-gated data) was the
strongest predictor of cardiac mortality, whereas the
severity and extent of ischemia (based on ungated
data) were the strongest predictor of future myocardial infarction [18]. Both studies show the prognostic
value of poststress ejection fraction rather than resting ejection fraction, indicating the importance of the
poststress study in defining ischemic dysfunction
(myocardial stunning). In many respects these are
not new data; it simply reflects the results of older
studies using exercise radionuclide ventriculography
[19,20].
Attenuation correction in SPECT imaging

Attenuation artifacts are a consequence of interactions between the gamma photons emitted by the
radioactive tracer within the myocardium and the
patients own tissues. They may be gender-specific
giving rise to predictable patterns of attenuation
(ie, breast in women and diaphragm in men) or related
to patient habitus, giving rise to unpredictable soft
tissue attenuation in different individuals. Attenuation
correction must be patient-specific, with an attenuation map created for each image acquisition. In practice, this is done using transmission tomography to
create an attenuation map much as CT uses transmission tomography to create an image. Various external
sources of radiation can be used but most commonly
gadolinium 153 (gamma radiation energy 100 KeV
and half-life 273 days) and an X-ray tube itself are
627
chosen [21]. Both have energies similar to 99mTc

(140 KeV) and 201Tl (67 KeV) and the resulting
transmission data can be used to correct for attenuation errors for perfusion images resulting from either
tracer. The transmission data set can be obtained
simultaneously with emission SPECT or separately.
The latter has a time penalty but the former needs
specific measures to prevent cross-talk between radiation from emission and transmission sources.
Image reconstruction is complex and beyond the
scope of this article. In lay terms, emission SPECT
uses filtered back-projection to create an image. To
use the transmission attenuation information, a variety of iterative reconstruction methods have been
developed. There is no consensus as to the best
approach and, as always in this situation, there are
many solutions. All manufacturers market different
hardware and software configurations for addressing
attenuation correction but the revolution in positron
emission tomography (PET) and CT for oncology
means CT attenuation correction is likely to prevail.
Does attenuation correction add value?
Early reports were encouraging with improved
specificity and reporter confidence compared with
uncorrected images [22 24]. Some authors go so
far as to argue that a normal attenuation-corrected
stress study obviates the need for a rest study. Several
studies, however, used protocols that involved more
than just simple attenuation correction (ie, additional
software correction for reducing depth-dependent
blur or scatter artifact). Although specificity is improved, sensitivity may be lower with new artifacts
occurring as a consequence of attenuation correction.
The unmasking of scatter artifacts is a particular
problem. When images are displayed on a scale
normalized to the region of highest activity, areas of
normal perfusion appear suppressed if the hot spot
has been overcorrected. As a result, overcorrection of
the inferior wall from subdiaphragmatic activity may
lead to an artifactual reduction in the anterior wall
and septum [25,26]. Other problem areas include
truncation artifacts, effects of patient and respiratory
motion, and an increase in apparent apical thinning.
Wackers [27] has recently proposed testing all
systems that use attenuation correction with two
standardized phantoms. This stems in part from a
report in which eight different attenuation compensation devices were compared [28]. The authors
showed that, under certain conditions, the application
of attenuation correction actually produced a decrease
in image quality and homogeneity, especially when a
simple cardiac phantom suspended in air was used.
628
There was a statistically significant improvement in

uniformity following correction when the cardiac
phantom was imaged within a second anthropomorphic phantom mimicking the chest wall. The
difference, however, was slight and was minimized
further by the presence of extracardiac activity especially in the liver.
The place of attenuation compensation in routine
clinical practice remains unclear. A recent position
statement from the American Society of Nuclear
Cardiology takes the stance that . . .attenuation correction has become a method for which the weight
of evidence and opinion is in favour of its usefulness
[29]. In contrast a consensus statement from a number
of British cardiac and nuclear societies takes a more
cautious view stating that these techniques . . .should
be used only in experienced centres and preferably as
part of a formal evaluation of their value [24,30].
Given the additional radiation burden and time penalty of attenuation correction, the authors believe it
cannot yet be justified in routine clinical practice.
Viability imaging with myocardial perfusion

single-photon emission CT and fluorodeoxyglucose
positron emission tomography
Viable myocardium may be normal, stunned, or
hibernating. Nonviable myocardium represents myocardial infarction or scar. Myocardial stunning refers
to dysfunctional myocardium where wall motion
abnormalities temporarily persist following a period
of ischemia, although normal blood flow is restored
[31]. Hibernating myocardium also is dysfunctional
but is caused by chronic down-regulation of contractile function in response to reduced perfusion reserve
[32]. The concept of hibernation has moved on since
the term was first used 15 years ago [33]. It is no
longer believed to be simply a low-flow phenomenon, because over half of reversibly dysfunctional
segments demonstrate normal transmural perfusion at
rest. Although resting flow may be normal, flow
reserve is always abnormal. Hibernation is the result
of repeated stunning rather than a separate cellular
state [34,35]. Stunning, hibernation, and cell death
represent a continuum of changes that, left untreated,
lead slowly to cellular dedifferentiation, degeneration, and myocardial fibrosis [36,37].
Differentiating these phases is important. Revascularizing dysfunctional scar confers no benefit,
whereas the revascularization of hibernating myocardium may improve left ventricular function. Whether
or not this translates into symptomatic benefit for the
patient depends on the volume of hibernating muscle
that is revascularized and can recover. Patients with

triple-vessel disease and poor left ventricular function
(traditionally a group with the highest surgical mortality) are also those who benefit most from coronary
artery bypass grafting. In a recent meta-analysis of
24 studies involving more than 3000 patients, there
was an 80% relative (13% absolute) risk reduction in
annual cardiac mortality following revascularization
of hibernating myocardium [38]. In patients with
predominantly nonviable dysfunctional myocardium
there was no benefit.
The question is how to distinguish dysfunctional
myocardium caused by hibernation, which may be
recoverable, from postinfarction scar, which is not
recoverable. If both hibernation and scar are present in
the same region, to what extent is recovery predictable? Many techniques have been advocated. In addition to the techniques proposed for nuclear medicine,
stress echocardiography, dobutamine stress MR imaging, and most recently MR perfusion with late gadolinium enhancement have been described [39,40].
Thallium 201 and technetium 99m sestamibi in
assessing myocardial viability
Thallium 201 acts as a marker of viability because
of its dependence on an intact cell membrane and
active transport for uptake. Because 201Tl redistributes freely after injection, viability may be assessed
from resting and redistribution images. A defect on
stress imaging that improves on delayed rest imaging
at 3 to 6 hours is viable tissue. About 50% of fixed
defects, however, are falsely characterized as scar
[41]. Higher sensitivity and specificity may be
obtained by imaging after reinjection [42] or delayed
imaging (24 hours after initial injection) [43]. Low
count statistics and long imaging times make late
imaging impractical for most centers and reinjection
confers a high radiation dose. Sestamibi uptake
depends on both perfusion and mitochondrial function and may also be used as a viability agent. Areas
of mismatch on stress and rest imaging represent
viable myocardium but simple reversible ischemia
or stunning cannot be distinguished from hibernation.
ECG-gating adds a functional perspective to sestamibi SPECT but contributes little to the detection of
dysfunctional but viable myocardium [44]. Most
accounts suggest that sestamibi is less accurate than
201
Tl for this purpose [45].
Whichever perfusion agent is used, there is a
direct correlation between the uptake of radioisotope
and tissue viability [46,47]. More importantly, the
extent of the uptake correlates with the likelihood of
segmental recovery following revascularization. For
an individual patient, recovery of one or two segments often makes little difference to either symptoms or prognosis. An increase in ejection fraction
predicts improved survival but to achieve this a significant volume of hibernating tissue needs to be
revascularized [48].
The combination of scar and hibernation in the
same poorly contracting segment of myocardium is a
629
particular problem (eg, non Q wave infarction). In

this situation there may be sufficient remaining viable
myocardium for tracer uptake but not enough for
functional recovery following revascularization. The
relatively poor spatial resolution of nuclear perfusion
makes the detection of subendocardial scar a problem
such that alternative techniques with better resolution
are needed.
Fig. 9. Bulls-eye plots comparing rest (A) and stress (B) sestamibi SPECT and FDG PET (C) in a patient with an established
infarct and hibernating myocardium. SPECT shows a very extensive severe fixed anteroapical defect with a moderately severe
fixed inferior defect. FDG PET shows the inferior defect to be active metabolically and likely to represent hibernating
myocardium. The inferior wall recovered function following revascularization.
630
Positron emission tomography and viability imaging

Assessment of viability by PET involves regional
comparison of myocardial perfusion with glucose
use. Rubidium 82, nitrogen 13 ammonia, and oxygen
15 water have all been used to assess and quantify
perfusion. The methods are technically difficult and
short tracer half-lives (seconds to minutes) mean they
are restricted to research centers sited close to a
cyclotron. Assessment of glucose use is another
matter. Fluorine 18 (18F) labeled fluorodeoxyglucose (FDG) has a relatively long half-life (110 minutes) and can be used at centers within a 2-hour
journey time of the cyclotron. FDG is taken up by
metabolically active myocytes and phosphorylated to
FDG-6-phosphate. The reverse reaction is virtually
absent and FDG accumulates in viable cells. In broad
terms, uptake in any one region is proportional to the
amount of viable myocardium in that region.
Normal myocardium is characterized by normal
perfusion and glucose uptake, whereas infarcted
myocardium by decreased perfusion and glucose
uptake. The most specific pattern for hibernating
myocardium is decreased perfusion with normal
glucose uptake. This mismatch pattern has a high
positive predictive value (ie, correctly identifying
those segments that show functional recovery following revascularization) and a high negative predictive
value (81% and 84%, respectively) (Fig. 9) [49].
Several studies have compared perfusion PET with
perfusion SPECT [50,51]. In general, sensitivities
were very similar, although one study showed a much
higher specificity for PET [52].
In the fasting state, the heart predominantly uses
free fatty acids as a source of fuel, whereas after a
meal a high circulating level of insulin promotes
glucose uptake. Glucose uptake is dependent on a
number of variables including dietary state, insulin
resistance, cardiac workload, sympathetic tone, and
ongoing ischemia. The response to a glucose load is
often unpredictable, particularly in type 2 diabetics
with insulin resistance, resulting in poor image quality. Standardized metabolic conditions are needed
if FDG is to be used reliably as a tracer. A number
of methods have been described [53], although the
authors and others have found the use of a hyperinsulinemic euglycemic clamp to be the most robust
[54,55].
Given the wide applicability of 18F PET compared
with the short half-life tracers used for PET perfusion
imaging (rubidium 82, nitrogen 13 ammonia, and
oxygen 15 water), the combination of FDG PET
and 99mTc SPECT may prove to be a good alternative
for detecting hibernation [56]. As with the all-PET
techniques, improvement in regional function following revascularization is directly proportional to the

mismatch of perfusion and metabolism (diminished
perfusion on SPECT imaging with relatively normal
glucose uptake by FDG PET) [57]. As the use of
FDG and the hardware necessary for oncology PET
becomes more widespread, it is likely that this
combined technique will become more popular.
Cost-effectiveness of nuclear perfusion imaging

Over the last 20 years countless papers have
been published describing the prognostic value of nuclear perfusion imaging. This is independent of clinical history, exercise-tolerance testing, and coronary
angiography. It is not surprising that many centers
are looking at its use in selecting patients for coronary angiography.
Nuclear perfusion imaging and the selection of
patients for angiography
For most patients, the diagnosis of chronic stable
angina is based on a combination of history and
exercise-tolerance testing. Those who are poorly
controlled on medication and could benefit from
revascularization are referred for angiography. Under
most circumstances, implicit in the referral for angiography, there is a referral for revascularization
should there be significant disease (ie, 70% diameter stenosis in a major coronary vessel, or 50% in
left main stem). The decision as to whether and how
to revascularize depends on the presence and distribution of significant disease. Most clinicians refer
single-vessel disease for percutaneous coronary intervention and triple-vessel disease for surgery. Debate
remains over the management of two-vessel disease
and this depends on vascular anatomy and local
expertise. In general, the presence of significant
disease leads to intervention. Only when a stenosed
or occluded vessel subtends nonfunctioning or infarcted myocardium is this likely to be left untreated.
Despite this apparently rational approach, there is
growing evidence that revascularization is not being
targeted appropriately. Comparing intervention in the
United States and Canada in patients with acute
coronary syndromes, rates of angiography, percutaneous coronary intervention, and coronary artery
bypass grafting within 6 months of presentation were
substantially lower in Canada than the United States
(69% versus 39%, 24% versus 13%, 25% versus
15%, respectively) [58]. The United States benefited
from a lower rate of refractory angina (14% versus
20%) but there was no difference in the rates of

myocardial infarction or cardiac death. Indeed, major
bleeding and stroke were significantly higher in the
United States. Reported rates of percutaneous coronary intervention and coronary artery bypass grafting relative to angiography are similar in the two
countries and are similar to those seen in the United
Kingdom (approximately 50% of patients undergoing
coronary angiography go on to revascularization).
Angiographic criteria for revascularization must be
internationally accepted. Because angiography is the
trigger for revascularization, variations in national
revascularization rates are most likely caused by
varying thresholds for angiography.
Why does primary angiography lead to a higher
rate of revascularization? The problem revolves
around defining the severity of intermediate stenoses
(ie, stenoses ranging from 40% to 70%). It is well
established that the impact of a stenosis on coronary
flow reserve is more important than its absolute
dimensions but determining a lesions functional
significance in vivo can be difficult. Perfusion imaging is a surrogate for functional significance but
cardiologists have been slow to accept it. It is hoped
that correlation of perfusion imaging with more
objective evidence from intracoronary Doppler wires
and derived measures of coronary flow reserve may
help [59]. Early work suggests that stress 201Tl
perfusion defects correlate significantly with those
coronary lesions that produce a significant reduction
in flow reserve [60]. More data are awaited to see if
the same is true for other agents in larger trials.
The negative predictive value of a normal or
equivocal perfusion study is well established, with
0% to 0.5% of patients going on to have a firm cardiac event (acute coronary syndrome, acute myocardial infarction, or cardiac death) in the next year
[61,62]. Even in the presence of an apparently highrisk exercise-tolerance test, the cardiac event rate
rises to only 0.7%. This is a similar level of cardiac
risk to age- and sex-matched controls with no evidence of cardiac disease. The converse is also true: an
abnormal perfusion study predicts a cardiac event rate
of 2.8% to 7.9%, the lower rate corresponding to a
low-risk exercise-tolerance test result and vice versa.
With results like these it is reasonable to consider
using nuclear perfusion imaging as a gatekeeper to
coronary angiography. Anticipating that the likely fall
in catheter numbers is mirrored by a fall in the
revascularization procedures, the strategy should be
cost neutral.
In a study comparing two risk-matched populations (one having angiography according to standard
criteria, the other having angiography only when
631
SPECT imaging was positive), revascularization was

more frequent in the primary angiography group
(51% compared with 38%) [63]. The ratio of revascularization to angiography was similar in both
groups indicating less angiography means less revascularization. As might be expected, a higher revascularization rate is reflected in higher costs, which
more than offsets the increased cost of functional
testing. In a similar study, Shaw et al [64] showed
that limiting angiography to patients with a positive
SPECT study reduced costs in diagnosis, revascularization, and follow-up. In both studies cardiac death
and myocardial infarction rates were unchanged despite lower revascularization rates in the SPECT
group. By refining selection criteria for angiography,
it may be possible to reduce both the rate of angiography and the rate of revascularization. This brings
not only cost benefits but also reduces patient risk
from unwarranted angiography and subsequent revascularization [65].
Summary
Nuclear medicine continues to evolve quickly
with new techniques, such as ECG-gating and attenuation correction helping to improve specificity and
reporter confidence. These improvements in accuracy have been matched by renewed interest in the
use of nuclear perfusion imaging as a gatekeeper
for coronary angiography. The scene is set for a
significant increase in demand for nuclear imaging.
Radiologists and imaging cardiologists must be prepared to meet the challenge. New methods of working are needed to increase efficiency and more
systematic reporting is essential if published standards of accuracy are to be met in nonspecialist
centers. The latter is particularly important if nuclear
perfusion imaging is to deliver the improvements in
patient selection for coronary angiography and revascularization that are anticipated.
References
[1] Coyne E, Belvedere D, Vande Streer P. Thallium-201
scintigraphy after intravenous infusion of adenosine
compared with exercise thallium testing in the diagnosis of coronary artery disease. J Am Coll Cardiol 1991;
17:1289 94.
[2] White C, Stone DL, Coulden RA. Does pharmacological stress in myocardial perfusion imaging need medical supervision? Nucl Med Commun 2000;21:395.
632
[3] Beharry N, White C, Stone DL, et al. Non-medical

supervision of pharmacological stress myocardial perfusion imaging. Nucl Med Commun 2003;24:335 6.
[4] Wackers FJT, Berman DS, Maddahi J. Technetium-99m
hexacis 2-methoxyosobutyl isonitrile: human biodistribution, dosimetry, safety and preliminary comparison to
thallium-201 for myocardial perfusion imaging. J Nucl
Med 1989;30:1989 95.
[5] Gibson P, Demus D, Noto R, Hudson B, Johnson L.
Low event rate for stress-only perfusion imaging in
patients evaluated for chest pain. J Am Coll Cardiol
2002;39:999 1004.
[6] Santana C, Garcia E, Vansant E, Krawczynska E, Folks
R, Cooke C, et al. Gated stress-only technetium 99m
myocardial perfusion SPECT imaging accurately
assesses coronary artery disease. Nucl Med Commun
2003;24:241 9.
[7] Candell-Riera J, Santana-Boado C, Bermejo B. Interhospital observer agreement in interpretation of exercise myocardial Tc-99m-tetrofosmin SPECT studies.
J Nucl Cardiol 2001;8:49 57.
[8] Cerqueira M, Weissman N, Dilsizian V. Standardized
myocardial segmentation and nomenclature for tomographic imaging of the heart. Circulation 2002;105:
539 42.
[9] Weiss A, Maddahi J, Lew A. Reverse redistribution of
thallium-201: a sign of non-transmural myocardial infarction with patency of the infarct-related coronary
artery. J Am Coll Cardiol 1986;7:61 7.
[10] Araujo W, DePuey E, Kamran M. Artefactual reverse
redistribution pattern in myocardial perfusion SPECT
with technetium-99m sestamibi. J Nucl Cardiol 2000;
7:633 8.
[11] DePuey E. How to detect and avoid myocardial perfusion SPECT artefacts. J Nucl Med 1994;53:699 702.
[12] Lette J, Lapoint J, Waters D, et al. Transient left ventricular cavity dilatation during diprydamole-thallium
imaging as an indicator of severe coronary disease. Am
J Cardiol 1990;66:1163 9.
[13] Germano G, Kiat H, Kavanagh PB. Automatic quantification of ejection fraction from gated myocardial perfusion SPECT. J Nucl Med 1995;36:2138 47.
[14] Taillefer R, DePuey E, Udelson J. Comparative diagnostic accuracy of Tl-201 and Tc-99m sestamibi
SPECT imaging (perfusion and ECG-gated SPECT)
in detecting coronary artery disease in women. J Am
Coll Cardiol 1997;29:69 77.
[15] Choi J, Lee K, Kim S. Gating provides improved accuracy for differentiating artefacts from true lesions in
equivocal fixed defects on technetium 99m tetrofosmin
perfusion SPECT. J Nucl Cardiol 1998;5:395 401.
[16] Smanio P, Watson D, Segalla D. Value of gating of
technetium 99m sestamibi single-photon emission
computed tomographic imaging. J Am Coll Cardiol
1997;30:1687 92.
[17] Sharir T, Germano G, Kavanagh P. Incremental prognostic value of post-stress left ventricular ejection fraction and volume by gated myocardial perfusion single
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
photon emission computed tomography. Circulation

1999;100:1035 42.
Sharir T, Germano G, Kang X. Prediction of myocardial infarction versus cardiac death by gated myocardial perfusion SPECT: risk stratification by the amount
of stress-induced ischaemia and the post-stress ejection
fraction. J Nucl Med 2001;42:831 7.
Jones R, Johnson S, Bigelow C. Exercise radionuclide
angiography predicts cardiac death in patients with
coronary artery disease. Circulation 1991;84:152 8.
Morris K, Palmeri S, Califf R. Value of radionuclide
angiography for predicting specific cardiac events after
acute myocardial infarction. Am J Cardiol 1985;55:
318 24.
Lang TF, Hasegawa BH, Liew SC, et al. Description of
emission-transmission computed tomography imaging
system. J Nucl Med 1992;33:1881 6.
Ficaro E, Fessler J, Shreve P. Simultaneous transmission/emission myocardial perfusion tomography:
diagnostic accuracy of attenuation-corrected 99mtechnetium sestamibi single-photon emission computed
tomography. Circulation 1996;93:463 73.
Hendel R, Berman D, Cullom S. Multicentre clinical
trial to evaluate the efficacy of correction for photon
attenuation and scatter in SPECT myocardial perfusion
imaging. Circulation 1999;99:2742 9.
Links J, Becker L, Rigo P, Taillefer R, Hanelin
L, Anstett F. Combined corrections for attenuation,
depth-dependent blur, and motion in cardiac SPECT:
a multicentre trial. J Nucl Cardiol 2000;7:414 25.
King MA, Tsui BMW, Pan TS. Attenuation compensation for cardiac SPECT imaging: Part 1. Impact of
attenuation and methods estimating attenuation maps.
J Nucl Cardiol 1995;2:513 20.
King MA, Tsui BMW, Pan TS. Attenuation compensation for cardiac SPECT imaging: Part 2. Attenuation compensation algorithms. J Nucl Cardiol 1995;
3:55 65.
Wackers F. Attenuation compensation of cardiac
SPECT: a critical look at a confusing world. J Nucl
Cardiol 2002;9:438 40.
OConnor M, Kemp B. A multicentre evaluation of
commercial attenuation compensation techniques in
cardiac SPECT using phantom models. J Nucl Cardiol
2002;9:361 76.
Hendel R, Corbet J, Cullom S, et al. The value and
practice of attenuation correction for myocardial perfusion SPECT imaging: a joint position statement from
the American Society of Nuclear Cardiology and the
Society of Nuclear Medicine. J Nucl Cardiol 2002;9:
135 43.
Anagnostopoulous C, Harbinson M, Kelion A, et al.
Procedure guidelines for radionuclide myocardial perfusion imaging. Heart 2004;90:1 10.
Braunwald E, Kloner RA. The stunned myocardium:
prolonged post ischaemic ventricular dysfunction. Circulation 1982;66:1146 9.
Vanoverschelde JL, Wijns W, Borgers M, et al. Chronic
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
myocardial hibernation in humans: from bench to bedside. Circulation 1997;95:1961 9.

Rahimtoola S. The hibernating myocardium. Am Heart
J 1989;117:211 21.
Fallavollita J, Canty J. Differential 18-F-2-deoxyglucose uptake in viable dysfunctional myocardium
with normal resting perfusion. Circulation 1999;99:
2798 805.
Shen Y, Vatner S. Mechanism of impaired myocardial
function during progressive coronary stenosis in conscious pigs: hibernation versus stunning? Circ Res
1995;76:479 88.
Ausma J, Schaart G, Thone F. Chronic ischaemic viable myocardium in man: aspects of dedifferentiation.
Cardiovasc Pathol 1995;4:29 37.
Schwarz E, Schoendube F, Kostin S. Prolonged
myocardial hibernation exacerbates cardiomyocyte degeneration and impairs recovery of function after revascularisation. J Am Coll Cardiol 1998;31:1018 26.
Allman K, Shaw L, Hachamovitch R. Myocardial viability testing and impact of revascularisation on prognosis in patients with coronary artery disease and left
ventricular dysfunction: a meta-analysis. J Am Coll
Cardiol 2002;39:1151 8.
Bax J, Wijns W, Cornel J. Accuracy of currently available techniques for prediction of functional recovery
after revascularisation in patients with left ventricular
dysfunction due to chronic coronary artery disease:
comparison of pooled data. J Am Coll Cardiol 1997;
30:1451 60.
Kim R, Wu E, Rafael A. The use of contrast-enhanced
magnetic resonance imaging to identify reversible
myocardial dysfunction. N Engl J Med 2000;343:
1445 53.
Jain D, Zraet BL. Nuclear imaging techniques for the
assessment of myocardial viability. Cardiol Clin 1995;
13:43 58.
Dilsizian V, Rocco T, Freedman N. Enhanced detection
of ischaemic but viable myocardium by the reinjection
of thallium after stress-redistribution imaging. N Engl J
Med 1990;323:141 6.
Kiat H, Berman D, Maddahi J. Late reversibility of
tomographic myocardial thallium-201 defects: an accurate marker of myocardial viability. J Am Coll
Cardiol 1988;12:1456 63.
Gonzalez P, Massardo T, Munoz A, et al. Is the addition of ECG gating to technetium-99m sestamibi
SPECT of value in the assessment of myocardial viability? An evaluation based on 2 dimensional echocardiography following re-vascularisation. Eur J Nucl
Med 1996;23:1315 21.
Marcassa C, Galli M, Cuocolo A, et al. Rest redistribution thallium-201 rest technetium-99m sestamibi
SPECT in patients with stable coronary artery disease
and ventricular dysfunction. J Nucl Med 1997;38:
419 25.
Maes A, Borgers M, Flameng W. Assessment of myocardial viability in chronic coronary artery disease
using technetium-99m sestamibi SPECT: correlation
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
633
with histologic and positron emission tomographic

studies and functional follow-up. J Am Coll Cardiol
1997;29:62 8.
Medrano R, Lowry R, Young J. Assessment of myocardial viability with 99m Tc sestamibi in patients
undergoing cardiac transplantation: a scintigraphic/
pathological study. Circulation 1996;94:1010 7.
Ragosta M, Beller G, Watson D. Quantitative planar
rest-redistribution 201Tl imaging in detection of myocardial viability and prediction of improvement in left
ventricular function after coronary artery bypass in
patients with severely depressed left ventricular function. Circulation 1993;87:1630 41.
Shelbert HR. Merits and limitations to viability and
future developments. J Nucl Cardiol 1994;1:S86.
Tamaki N, Yonekura Y, Senda M, et al. Value and
limitation of stress thallium-201 single photon emission computed tomography: comparison with nitrogen13 ammonia positron tomography. J Nucl Med 1988;
29:1181 8.
Go R, Marwick T, MacIntyre W. A prospective comparison of rubidium-82 PET and thallium-201 SPECT
myocardial perfusion imaging utilising a single dipyridamole stress in the diagnosis of coronary artery
disease. J Nucl Med 1990;31:1899 905.
Stewart R, Schwaiger M, Molina E. Comparison of
rubidium-82 positron emission tomography and thallium-201 SPECT imaging for the detection of coronary
artery disease. Am J Cardiol 1991;67:1303 10.
Bacharach S, Bax J, Case J, Delbeke D, Kurdziel K,
Martin W, et al. PET myocardial glucose metabolism
and perfusion imaging: American Society of Nuclear
Cardiology Practice Guidelines. J Nucl Cardiol 2003;
10:543 54.
DeFronzo R, Tobin J, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and
resistance. Am J Physiol 1979;237:E214 23.
Dutka D, Camici P. The contribution of positron emission tomography to the study of ischaemic heart failure. Prog Cardiovasc Dis 2001;43:399 418.
Stoll HP, Hellwig N, Alexander C, et al. Myocardial
metabolism imaging by means of fluorine-18 deoxyglucose/technetium-99m sestamibi dual isotope singlephoton emission tomography. Eur J Nucl Med 1994;
21:1085 92.
Di Carli M, Asgarzadie F, Schelbert H, et al. Quantitative relation between myocardial viability and
improvement in heart failure symptoms after revascularisation in patients with ischaemic cardiomyopathy.
Circulation 1995;92:3436 44.
Yusuf S, Flather M, Pogue J, et al. Variations between
countries in invasive cardiac procedures and outcomes
in patients with suspected unstable angina or myocardial infarction without initial ST elevation. Lancet
1998;352:507 14.
Aude Y, Garza L. How to prevent unnecessary coronary interventions: identify lesions responsible for
ischaemia in the cath lab. Curr Opin Cardiol 2003;18:
394 9.
634
[60] Yanagisawa H, Chikamori T, Tanaka N, et al. Correlation between thallium-201 myocardial perfusion defects and the functional severity of coronary artery
stenosis as assessed by pressure-derived myocardial
fractional flow reserve. Circ J 2002;66:1105 9.
[61] Groutars RG, Versijlbergen JF, Muller AJ, et al. Prognostic value and quality of life in patients with normal
rest thallium-201/stress technetium-99m tetrofosmin
dual isotope myocardial SPECT. J Nucl Cardiol
2000;7:333 41.
[62] Berman D, Hachamovitch R, Kiat H, et al. Incremental
value of prognostic testing in patients with known or
suspected ischaemic heart disease: a basis for optimal
utilisation of exercise technetium-99m sestamibi myocardial perfusion single photon emission computed
tomography. J Am Coll Cardiol 1995;26:639 47.
[63] Mishra JP, Acio E, Heo J, et al. Impact of stress single

photon emission computed tomography perfusion
imaging on down stream resource allocation. Am J
Cardiol 1999;83:1401 3.
[64] Shaw LJ, Hachamovitch R, Berman DS, et al. The economic consequences of available diagnostic strategies
for the evaluation of stable angina patients: an observational assessment of the value of precatheterization ischaemia. J Am Coll Cardiol 1999;33:661 9.
[65] Patterson R, Eisner R, Horowitz S. Comparison of
cost-effectiveness and utility of exercise ECG, single
photon emission computed tomoraphy, positron emission tomography and coronary angiography for the
diagnosis of coronary artery disease. Circulation
1995;91:54 65.
Radiol Clin N Am 42 (2004) 635 649
Multidetector-row CT of the heart

U. Joseph Schoepf, MDa,*, Christoph R. Becker, MDb,
Lars K. Hofmann, MDc, E. Kent Yucel, MDd
a
Department of Radiology, Medical University of South Carolina, 169 Ashley Avenue, Charleston, SC 29425, USA
Institute of Clinical Radiology, University of Munich, Klinikum Grosshadern, Marchioninistrasse 15, 81377 Munich, Germany
c
Siemens Medical Solutions, Division CT, Siemensstrase 1, 91301 Forchheim, Germany
d
Department of Radiology, Brigham and Womens Hospital, Harvard Medical School, 75 Francis Street, Boston, MA 02115, USA
b
Coronary artery disease (CAD) remains the leading cause of death in western nations. The standard of
reference for diagnosis of CAD is coronary catheter
angiography. In the year 1999 more than 1.83 million
cardiac catheter examinations were performed in the
United States [1]. The greatest advantage of catheter
angiography is its high spatial resolution and the
option of directly performing interventions, such as
balloon dilatation or coronary stenting. Only one third
of all coronary catheter examinations in the United
States were performed in conjunction with an interventional procedure (percutaneous transluminal coronary angioplasty [PTCA]), however, whereas the rest
were performed for mere diagnostic purposes (ie, for
verifying the presence and degree of CAD only) [1].
Accordingly, a reliable, noninvasive tool for imaging
of the coronary arteries and for early diagnosis of
CAD is highly desirable.
Imaging of the heart has always been technically
challenging, because of the hearts continuous
motion. CT imaging of the heart moved into the
diagnostic realm by the introduction of electron-beam
CT [2] and multidetector-row CT (MDCT) [3,4] and
the development of ECG-synchronized scanning and
reconstruction techniques [5]. These modalities allow
for faster volume coverage and higher spatial and
This article is being reprinted from the May 2003

issue of the Radiologic Clinics of North America, pages
491 505.
E-mail address: schoepf@bwh.harvard.edu
(U.J. Schoepf).
temporal resolution. The introduction of MDCT

especially has greatly benefited cardiovascular CT
imaging applications. The speed of image acquisition
shortens breathhold and examination times for the
patient and reduces the amount of contrast media
needed for high and consistent vascular enhancement
[5 9]. With the advent of 16-slice MDCT scanners
submillimeter resolution of substantial anatomic volumes is routinely achieved [8].
Until recently, cardiac CT applications were almost exclusively directed at the detection and quantification of coronary calcium. It is increasingly
recognized, however, that the diagnostic value of
CT coronary calcium measurements alone is limited.
Contrast-enhanced MDCT may effectively address
these limitations. Investigation of the heart with
submillimeter spatial resolution and a temporal resolution of 200 milliseconds now enables accurate,
high-resolution morphologic evaluation of both the
myocardium and the coronary arteries. Because of the
cross-sectional nature of CT, the vessel wall can be
evaluated. The potential of this technique for noninvasive identification of vulnerable atherosclerotic
lesions, which may have a higher propensity to
trigger acute coronary syndromes than stable, calcified plaques, is an area of active research. MDCT
may become a valuable tool for noninvasive atherosclerosis imaging.
MDCT image acquisition

In 1998, MDCT systems with four detector arrays
and a minimum rotation time of 500 milliseconds
doi:10.1016/j.rcl.2004.03.011
636
U.J. Schoepf et al / Radiol Clin N Am 42 (2004) 635649
were introduced [3,4,10] that provide an up to eightfold performance compared with 1-second rotation
single-slice CT systems. The performance of current
16-slice CT systems with 420-millisecond rotation
now represents a nearly 40-fold improvement as
compared with the acquisition speed of single-slice
CT [8]. The combination of fast rotation time and
multidetector-row acquisition became of particular
importance for cardiac applications.
To reduce motion artifacts caused by cardiac
pulsation, it is necessary either to scan or reconstruct
raw data at a time point with the least cardiac motion
(ie, in the diastole of the heart cycle). For successful
electrocardiogram (ECG) synchronization, prospective ECG triggering and retrospective ECG gating are
the two strategies that are used most commonly.
Prospective ECG triggering has long been used in
conjunction with electron beam CT and more recently
with single-slice spiral CT [11 13]. A prospective
trigger signal is derived from the patients ECG and
the scan is started at a defined time point, usually
during diastole. MDCT allows simultaneous acquisition of several slices within one heartbeat. The shorter
scan times facilitate routine clinical application. This
technique is also the most dose-efficient way of ECG
synchronization [7]. Usually, however, a rather thick
collimation (3 mm with electron beam CT, 2.5 mm
with four-slice CT) is used for prospectively triggered
acquisition. Resulting data sets are not as suitable for

three-dimensional reconstruction as thin collimation,
retrospectively gated scan data. Also, prospectively
ECG-triggered technique greatly depends on a regular heart rate of the patient and is bound to result in
misregistration in the presence of arrhythmia.
Retrospective ECG gating (Fig. 1) effectively
overcame the limitations of prospective ECG triggering with inconsistent heart phase scanning in patients
with arrhythmia. For this approach slow table motion
during spiral scanning and simultaneous acquisition
of currently 4, 8, or 16 slices and the digital ECG
trace are used to perform an oversampling of scan
projections [5,8]. Retrospective ECG gating creates
image stacks reconstructed at the exact same phase of
the heart cycle, which cover the entire volume of the
heart or great vessels (see Fig. 1). Usually the diastolic phase of the cardiac cycle is chosen for image
reconstruction for routine evaluation of cardiac morphology; however, because of data oversampling
during scan acquisition, image data is available for
each x, y, and z position within the scanned volume
over the entire course of the cardiac cycle. In the
presence of arrhythmia, the reconstruction interval for
each individual image stack can be shifted arbitrarily
within the cardiac cycle, so that reconstruction always
coincides with the same interval during diastole at
each level of the cardiac volume.
Fig. 1. Graph shows reconstruction with retrospectively electrocardiogram (ECG)-gated four-slice multidetector-row CT
(MDCT) scanning. The same basic principle also applies to newer generations of 8-slice and 16-slice CT scanners. Oversampled
scan data and the ECG of the patient are simultaneously recorded. Based on input from both data sources retrospective ECG
gating creates image stacks reconstructed at the same phase of the heart cycle. Usually diastole is chosen to suppress cardiac
motion. In this manner the entire volume of the heart (z) is covered within one breathhold (Adapted from Ohnesorge B, Flohr T,
Becker C, et al. Cardiac imaging by means of electrocardiographically gated multisection spiral CT: initial experience.
Radiology 2000;217:564 71; with permission.).
A prerequisite for successful imaging of the

coronary arteries is sufficient acquisition speed to
suppress cardiac motion and high spatial resolution
for accurate visualization of small sized vessels.
The newest generation of MDCT scanners seems
the most promising modality to date for fulfilling
these requirements.
Even during diastole, an estimated temporal resolution of 250 milliseconds is needed for a heart rate
of up to 70 bpm for motion-free imaging. For a heart
rate of up to 100 bpm the temporal resolution needs to
be increased to 150 milliseconds. With new-generation MDCT and dedicated, optimized spiral reconstruction algorithms a temporal resolution of only
slightly more than 100 milliseconds per image [8] can
be achieved, depending on the heart rate of the patient.
Coronary arteries are small and complex threedimensional structures. The diameter of coronary
vessels tapers down from typically 4 mm in the left
main coronary artery to 1-mm luminal diameter in the
distal left anterior descending coronary artery. To
image these structures properly, in-plane and throughplane spatial resolutions of 1 mm and less are necessary. Newer-generation 16-slice CT scanners now
provide an effective through-plane (z) resolution of
0.6 mm, which greatly improves visualization of subtle
pathology along the tortuous course of the coronary
artery tree.
For detection and quantification of coronary artery
calcium usually non contrast-enhanced scan technique is used, although accurate measurement of
coronary calcium is also feasible based on a thinslice MDCT angiography protocol [14]. For intravenous coronary angiography contrast media injection
must be tailored carefully either by using a test bolus
or automatic bolus triggering technique. Because
scan times for imaging of the heart on modern 8- or
16-slice MDCT scanners range from 20 to 40 seconds, 80- to 120-mL contrast media at injection rates
between 3 and 5 mL/second is needed to maintain
homogenous vascular contrast throughout the scan.
Saline chasing has proved useful for reduction of
contrast media needed for high and consistent vascular enhancement and for avoiding streak artifacts,
which frequently arise from dense contrast material in
the superior vena cava and the right atrium and
sometimes interfere with the evaluation especially
of the right coronary artery. Techniques for contrast
bolus optimization have been developed in the past
[15 17] but have not been widely used because
reasonable results could be obtained by adapting
single-slice CT strategies for contrast administration
to dual- and 4-slice CT. The introduction of ever
faster CT acquisition techniques, however, now
637
requires careful custom tailoring of the bolus for

achieving adequate and consistent contrast media
attenuation within the cardiovascular system.
Data visualization
Visualization of high-resolution MDCT data sets
consisting of several hundred individual axial images
is a daunting task. Still, some findings, such as
atherosclerotic lesions within the vessel wall, are best
evaluated based on individual axial sections. For better
visualization of the coronary artery tree in its entirety,
the following strategies are most commonly used.
Maximum intensity projection
For visualization of the coronary artery tree at
contrast-enhanced MDCT coronary angiography,
maximum intensity projections (MIPs) [18] are a
robust and easy to perform secondary visualization
tool for data viewing in daily clinical practice. Using
MIPs or other two-dimensional or three-dimensional
visualization methods (see later discussion) for diagnosis not only displays coronary artery MDCT data in
a more intuitive format but also condenses diagnostic
information into few relevant sections or views if
appropriate strategies are chosen. For routine visualization of large-volume MDCT coronary angiography datasets, the authors routinely perform three
dedicated MIP reconstructions to create views of
the left (Fig. 2A) and right (Fig. 2B) coronary arteries
and of the entire coronary arterial tree from a craniooblique perspective (spiderview) (Fig. 2C).
Multiplanar reformats
Another simple tool for secondary visualization of
high-resolution MDCT coronary angiography data is
use of multiplanar reformats. Because of the near
isotropic nature (equal voxel dimensions in x, y, and
z axis) of high-resolution MDCT acquisitions, image
data can be rearranged in arbitrary imaging planes
with comparable image quality as in the original axial
section. An additional option is creating curved
multiplanar reconstructions, which is especially useful to follow the course of coronary arteries (Fig. 3).
Three-dimensional visualization
Especially for nonradiologists it often is difficult
to mentally convert two-dimensional axial images
into three-dimensional anatomic information. Threedimensional postprocessing is a means to intuitively
638
Fig. 2. Maximum intensity projections (MIPs) are routinely used for the display of the coronary artery tree at MDCT coronary
angiography and are a robust and easy to perform secondary visualization tool for data viewing in daily clinical practice. For
routine visualization of large-volume MDCT coronary angiography datasets the authors routinely perform three MIP reconstructions. Views are created of the left (A, note spotty calcifications of the left anterior descending coronary artery) and right
(B) coronary arteries and of the entire coronary arterial tree from a cranio-oblique perspective (C, note calcification of the left
anterior descending coronary artery [arrow]).
639
and soft tissue analysis could be complemented with

the superior temporal and spatial resolution of MDCT
data (Fig. 6).
Clinical applications
Calcium scoring: clinical rationale
Advanced visualization tools
Because arterial calcification almost always represents atherosclerosis, detection of coronary artery
calcium by means of CT is a sensitive, noninvasive
tool for determining the presence of coronary atherosclerosis [19]. The absence of coronary calcification
at CT has a high negative predictive value for ruling
out the presence of atherosclerosis and of stenotic
CAD (eg, in a population of patients with atypical
chest pain) [20 22]. Attempts have been made to use
the presence and degree of coronary calcification for
determining the extent and location of stenotic disease [23 25] and for defining patients at risk of hard
cardiac events (ie, unstable angina, myocardial infarction, need for revascularization, coronary death)
[24,26]. Early excitement has been tempered, however, by the results of meta-analyses pooling prognostic data on the positive predictive value of an
elevated calcium score. According to these analyses
there is only a very moderately increased risk for hard
Advanced software tools are actively being developed that facilitate viewing and analysis of large
volume data sets. Dedicated software algorithms
allow for automated segmentation and extraction of
the coronary artery tree from contrast-enhanced CT
studies of the heart. Intuitive visualization of the entire
course of a coronary artery can be achieved by displaying a curved multiplanar reformat along an automatically generated centerline of the vessel (Fig. 5).
It needs to be determined whether such tools are able
to increase the accuracy for lesion detection and
stenosis quantification.
Similarly, efforts are being directed at image
coregistration from different image modalities, such
as MR imaging and CT. Cardiac MR imaging is used
successfully for analysis of myocardial function and
perfusion and allows assessing myocardial viability
by differentiating myocardial scars from areas of
hypoperfusion or hibernation. Combining structural
CT information on coronary artery lesions with
functional MR imaging information on the state of
myocardial motion, perfusion, and viability enables
gauging the functional significance of atherosclerotic
lesions for choice of adequate therapeutic regimens.
MR imaging advantages for functional assessment
Fig. 4. Colored volume rendering of a noninvasive MDCT

coronary arteriogram scanned with 16-slice MDCT and
420-millisecond rotation speeds allows visualization of the
right and left anterior descending coronary arteries. Extensive
atherosclerotic calcifications are noted along the course of the
right coronary artery. Three-dimensional postprocessing is a
means to intuitively display and convey information on the
often complicated anatomy of tortuous coronary arteries.
Fig. 3. Curved multiplanar reformat of a contrast-enhanced

MDCT coronary angiography study allows visualization of
the course of the left anterior descending coronary artery in a
patient with coronary artery disease. Note significant stenosis
(arrow) caused by a noncalcified coronary artery lesion
proximal to a calcified nodule.
display and convey information on the often complicated anatomy of tortuous coronary arteries. The most
commonly used technology for three-dimensional
visualization of the coronary arterial tree is volume
rendering (Fig. 4).
640
Fig. 5. Dedicated software platform for automated segmentation and extraction of the coronary artery tree from contrastenhanced CT studies of the heart (right upper image panel of user platform). Intuitive visualization of the entire course of the left
anterior descending coronary artery is achieved by displaying a curved multiplanar reformat along an automatically generated
centerline of the vessel (lower image panel of user platform).
cardiac events associated with coronary calcifications

detected at CT in high-risk, asymptomatic populations [27,28]. Also, the accuracy of coronary calcium
measurements for predicting coronary artery stenosis
as compared with cardiac catheterization seems to be
only very moderate [28]. According to prevailing
opinion, the total amount of coronary artery calcium
cannot be regarded as a direct predictor of hard
cardiac events and the incremental prognostic value
of coronary calcium compared with that of traditional
risk factor assessment remains to be defined fully.
Large prospective trials in the general population are
needed to define subgroups that might benefit from
quantitative assessment of coronary calcium. Current
more discriminating considerations on CT imaging of
coronary artery calcium mainly focus on the role of
calcium in the pathogenesis of atherosclerotic disease. Coronary atherosclerosis is a systemic disease
process. The presence and extent of coronary artery

calcifications may be considered indicative of the
total burden of calcified and noncalcified plaque of
a given individual and also of the likelihood of the
presence of potentially vulnerable coronary artery
lesions. Imaging of coronary calcium, although unable to identify a localized coronary artery lesion,
potentially may play a role in identifying the more
vulnerable patient. With this rationale, the degree of
coronary artery calcifications may be considered a
risk factor, and depending on the outcome of large
population-based studies currently underway, as such
may become part of the traditional Framingham risk
stratification scheme in the future. It can be foreseen
that the degree of coronary atherosclerosis, as determined based on total calcified plaque burden, may
replace age as an independent risk factor in the traditional risk stratification schemes.
641
Fig. 6. Prototype software platform enabling spatial image coregistration of MR imaging (right upper image panel of user interface)
and CT (left upper image panel of user interface) data. Combining structural CT information on coronary artery lesions with
functional MR imaging information on the state of myocardial motion, perfusion, and viability enables comprehensive assessment
of cardiac morphology and function.
Calcium scoring: technique

Imaging of coronary artery calcium has been
performed with electron-beam CT [23], single-slice
CT [12,13,29,30], dual-slice CT [31], and MDCT
[32 35]. A known limitation of coronary artery calcium scoring is the high interscan variability associated with this test [36]. This high variability has
limited the use of coronary artery calcium measurements for tracking the progression of atherosclerosis
under statin (lipid-lowering) therapy, which may
become a potentially powerful future application of
this technique [37,38]. The most promising technology to overcome this problem seems to be use
of MDCT technology with retrospective ECG gating.
Recent studies investigating this technique found an
interscan variability of only 5% at repeat MDCT
scanning [34,39], which may be accurate enough
to sensitively detect changes in the total atherosclerotic disease burden in patients with and without
specific therapy. As compared with prospectively
ECG-triggered technique, MDCT acquisition with
retrospective ECG gating is associated with higher effective radiation exposure of the patient (ie, 2 mSv in
men and 2.5 mSv in women [40]). Frequently, healthy,
asymptomatic individuals undergo coronary calcium
scoring in the context of primary prevention. Especially in this population it is imperative to keep radiation dose to a minimum. This can be achieved by
adapting scan protocols accordingly [41], or by using
sophisticated technical developments, such as ECGbased tube current modulation [40], which can
decrease effective radiation exposure of the patient
by as much as 50% [40].
The most commonly used algorithm for quantification of coronary artery calcium is the traditional semi-
642
quantitative score based on slice-by-slice analysis

of CT images as described by Agatston et al [23].
Recent studies describe better results for interscan
and interobserver and intraobserver variability with
use of a quantitative volume score as compared with
the traditional Agatston scoring method [34,39,42].
Advanced software platforms (Fig. 7) additionally
allow determining total calcified plaque burden in
terms of absolute calcium mass based on actual
scanner-specific calibration [39,43]. This latter technique probably has the greatest potential to increase
accuracy and reproducibility of coronary calcium
assessment [43] and with some likelihood will replace
traditional scoring methods in the future [44].
Cardiac function
Noninvasive assessment of cardiac function is
performed routinely with use of echocardiography.
In patients undergoing invasive cardiac catheteriza-
tion, left ventricular volumes and cardiac function

can be determined by levocardiography using monoplane or biplane projections. Both techniques are
mainly based on geometric assumptions for calculating left ventricular volumes. Cross-sectional imaging
allows for three-dimensional calculation of cardiac
volumes based on covering the ventricles with
contiguous slices along the intrinsic cardiac axis
[2,45,46]. Because of the slow data oversampling,
which is performed for contrast-enhanced MDCT of
the heart with retrospective ECG gating, cross-sectional images of the entire cardiac volume can be
reconstructed at any desired phase during the cardiac
cycle based on the patients ECG. Usually, image
stacks are only reconstructed during diastole to
freeze cardiac motion, whereas image data acquired
over the rest of the ECG goes unused. If so desired,
however, basic cardiac function parameters, such as
left and right ventricular ejection fraction and myocardial wall thickness, can be assessed by perform-
Fig. 7. Commercial semiautomated coronary calcium scoring software platform. Three-dimensional based selection and viewing
tools are used to identify calcified lesions and to attribute them to different vascular territories (left main, left anterior descending,
circumflex, right coronary artery). The most common algorithms for quantification of coronary artery calcium are the traditional
Agatston score, volume scores, and total calcium mass.
ing volume reconstructions during end systole and

end diastole [46]. Reformation of long and short axis
views are used to facilitate segmentation of ventricular volumes in different heart phases (Fig. 8).
Recent studies show good correlation between function parameters derived from MDCT and gold standard methods, such as levocardiography [46].
Functional information is inherently available in
retrospectively ECG-gated MDCT acquisitions (see
previous discussion) at no extra cost in terms of scan
time and radiation exposure to the patient and should
be used if so desired and if additional diagnostic
information can be obtained. Because this is not
the mainstay of CT, however, dedicated contrastenhanced MDCT merely aimed at function analysis
should be limited to patients who cannot be evaluated by less invasive methods, such as standard
echocardiography or MR imaging. Potential indications may include patients with emphysema or with
contraindications to MR imaging (metal implants,
pacemakers, and so forth).
643
MDCT coronary angiography: clinical rationale

The greatest challenge for noninvasive imaging is
reliable assessment of the coronary arteries, because
of their small size, tortuous three-dimensional anatomy, and fast continuous motion. Because of the
overwhelming importance of CAD in western economies, accurate noninvasive evaluation of coronary
arteries is a coveted goal. No noninvasive modality
has yet quite tackled this task. Intravenously contrastenhanced MDCT coronary angiography, however,
currently seems to fulfill the requirements best for
noninvasive morphologic assessment of the coronary
arteries, based on its unprecedented acquisition
speed, spatial resolution, and robustness of use.
Because of these features, current viable indications for MDCT coronary angiography already include assessment of anatomic anomalies of the
coronary arteries (Fig. 9) and evaluation of coronary
bypass graft patency (Fig. 10). A recent study demonstrated a sensitivity of 97% and a specificity of 89%
Fig. 8. MDCT evaluation of myocardial function. A retrospectively ECG gated data set is reconstructed during end-diastole
(upper left image panel of user interface) and end-systole (upper right image panel of user interface). Shown are multiplanar
reformats along the short axis of the heart. A dedicated software algorithm is used to quantify myocardial thickening during
systole within different sections of the myocardium on a color-coded map (lower left image panel of user interface).
644
Fig. 9. Catheter angiography (A) and volume rendered reconstruction of a contrast enhanced 16-slice CT coronary angiography
(B) in a patient with a super-dominant right coronary artery (RCA). The RCA gives rise to two major branches, which cross over
to the left anterior surface of the heart, connecting the RCA with the left anterior descending territory.
for detecting occlusions in bypass grafts [47]. Patency

of coronary artery stents can be determined with 98%
specificity with use of CT [48]. Limitations for accurate assessment of in-stent restenosis are related to
metal (blooming) artifacts that in some instances
compromise accurate visualization of the stent lumen
Fig. 10. Patient with left internal mammary artery bypass

graft (arrows). An anastomosis has been created between the
left internal thoracic artery and the left anterior descending
coronary artery territory. Note extensive atherosclerotic
changes in the native vessels.
[6,49,50]. Current generations of 16-slice CT scanners

with improved through-plane resolution may be able
to offset most of these limitations [8].
The accuracy of CT angiography for noninvasive
stenosis detection is an area of active research. Depending on study design, most published series using
four-slice CT technology found a sensitivity of noninvasive CT angiography for the detection of hemodynamically significant coronary artery stenosis
within proximal coronary arteries ranging between
80% and 90% [50 55]. Four-slice CT may not
suffice for reliably ruling out significant stenosis
in daily clinical routine. Most published studies
demonstrate a very high negative predictive value
of MDCT coronary angiography, however, so that the
absence of findings on contrast-enhanced CT angiography may be used to rule out the presence of significant CAD. The advent of faster MDCT scanner
generations with added detector elements is expected
to improve the overall accuracy of noninvasive CT
coronary angiography for stenosis detection (Fig. 11)
[56], which may decrease the number of invasive
coronary angiograms that are performed solely for
diagnostic purposes. The accuracy of MDCT for
stenosis detection may also be improved in the future
by the development of increasingly sophisticated
software tools aimed at facilitated visualization of
large-volume MDCT data sets (Figs. 5 and 12) and
eventually at automated detection of lesions causing
coronary artery stenosis (see Fig. 12).
Fig. 11. A 57-year-old female patient with high-grade

stenosis (arrow) of the proximal left anterior descending
coronary artery. Colored volume rendered display of a
contrast enhanced MDCT coronary angiogram.
MDCT coronary angiography: technique

The overall diagnostic quality of noninvasive
MDCT coronary angiography depends on many factors, among which choice of the appropriate reconstruction time point within the cardiac cycle, patient
heart rate, and contrast enhancement are of paramount importance. The motion pattern of the left
645
anterior descending and circumflex coronary arteries

(Fig. 13) follows the left heart, whereas the right
coronary artery (Fig. 14) moves synchronous with the
right heart. Because of these different motion patterns, different reconstruction time points over the
cardiac cycle may result in optimal display of different coronary arteries [57 59].
Most studies agree that patient heart rate is inversely related to diagnostic image quality at MDCT
coronary angiography [53,57,60 62]. It seems
recommendable (ie, oral administration of b-blockers)
to pharmacologically slow down the heart rate of
individuals undergoing MDCT coronary angiography
to less than 60 bpm with four-slice MDCT and less
than 70 bpm with 16-slice MDCT after contraindications to such a regimen have been ruled out.
Optimization of contrast media injection protocols
for MDCT coronary angiography is aimed at providing
homogenous enhancement within the entire course of
the coronary arteries to facilitate density-threshold
dependent two-dimensional and three-dimensional
visualization. Optimal contrast attenuation within the
vessel should be high enough to allow for lesion detection but not as high as to obscure calcified coronary
artery wall lesions with higher Hounsfield unit attenuation (ie, > 350 Hounsfield unit). With four-slice CT
(approximately 40-second scan time) this is achieved
in most patients with 140 mL of 300 mg/mL iodinated
contrast material injected at a flow rate of 3.5 mL/
second. Because of increased acquisition speed with
Fig. 12. Prototype software based on the clinical tool in Fig. 5. An automated centerline is created along the vessel lumen. The
lumenal profile is automatically analyzed and a stenosis caused by a partially calcified atheromatous lesion is detected by
the algorithm.
646
Fig. 13. Noninvasive MDCT coronary angiography. Volume

rendering of the left main coronary artery with its branches,
the left anterior descending, and left circumflex arteries in
anteroposterior cranial projection (right). Comparison with
conventional selective coronary angiography in the same
patient (left).
16-slice CT (approximately 20-second scan time) the

amount of contrast media can be reduced to 80 to
100 mL, delivered at an injection rate of 4 mL/second.
Use of saline-chasing technique (eg, with a bolus of
50 mL of saline injected immediately after the iodine
bolus) may be helpful for better contrast bolus use and
for reducing streak artifacts arising from dense contrast
material in the superior vena cava and the right heart.
MDCT imaging of the vulnerable plaque
An inherent advantage of MDCT for imaging of
the coronary arteries is the cross-sectional nature of
this technology. Conventional catheter angiography,
widely accepted as the gold standard for the detection
of CAD because of its unsurpassed spatial resolution,
displays only the vessel lumen and the degree of
lumenal narrowing in a cast-like manner, but fails to
visualize the coronary artery wall. In contrast, MDCT
enables detection of lesions within the coronary artery
Fig. 14. Noninvasive MDCT coronary angiography. Volume

rendering of the right coronary artery in 30-degree right
anterior oblique projection (right). Comparison with conventional selective coronary angiography in the same
patient (left).
wall that may or may not cause lumenal stenosis

(Figs. 3, 15) [9,63 67]. Another modality that has
this ability is intravascular ultrasound; however,
because of the invasive and expensive nature of this
test it seems unsuited for use in the context of primary
prevention. MR imaging also is capable of imaging
vessel wall structures and of differentiating various
stages of atherosclerotic lesions [68 71]. In vivo MR
imaging of the coronary artery wall, however, is
challenging because of a combination of cardiac and
respiratory motion artifacts, the tortuous course, small
size, and location of the vessels. Development of
more robust and reliable MR imaging technology is
needed before in vivo plaque assessment in human
coronary arteries becomes a viable clinical reality.
The ability of MDCT to noninvasively visualize
atherosclerotic processes within the vessel wall
[9,63 67] has sparked considerable scientific interest
and may provide more valuable insight in the intricate
pathogenesis of coronary atherosclerosis than
imaging of coronary calcium (see above). Coronary
calcium is intimately associated with coronary atherosclerotic plaque development but represents an
advanced stage of vascular remodeling in response
to atherosclerotic lesions [72]. Histopathologic studies
have demonstrated that calcium is a frequent feature of ruptured plaques (ie, culprit lesions associated
with acute coronary syndromes), but the presence
or absence of calcium does not allow for reliable
distinction between unstable versus stable plaques
[73 75]. Earlier, more active stages of coronary
Fig. 15. Curved multiplanar reformat of a contrast-enhanced

CT coronary angiography study in a 59-year-old man with
extensive coronary atherosclerosis. A complex atheromatous
lesion is visualized in the right coronary artery (arrow).
Adjacent to a calcified nodule the soft tissue component of
the lesion with lipid-like Hounsfield unit attenuation causes
significant stenosis of the vessel.
atherosclerosis seem more frequently associated with

noncalcified or mixed plaque composition, consisting
of accumulations of extracellular lipid and fibrous
tissue [76,77]. This may serve to explain the results
of clinical studies arguing that acute coronary syndromes occur more frequently in the absence of coronary calcium and that the presence of more extensive
calcification is more characteristic of stable CAD
[78]. Meanwhile, there is indication that morphology
and Hounsfield unit attenuation of coronary artery
lesions detected at contrast-enhanced, high-resolution
MDCT may allow noninvasive assessment of plaque
composition [65,66,71]. Future studies will show
to what extent this feature of MDCT, which is
currently actively researched, might translate into
the ability to detect preclinical atherosclerotic lesions,
stratify cardiac risk in vulnerable populations, and
monitor disease progression or lesion stabilization
under specific therapy.
References
[1] AHA. 2002 heart and stroke statistical update. Dallas
(TX): American Heart Association; 2001.
[2] Lipton MJ, Higgins CB, Boyd DP. Computed tomography of the heart: evaluation of anatomy and function.
J Am Coll Cardiol 1985;5:55S 69S.
[3] McCollough CH, Zink FE. Performance evaluation of
a multi-slice CT system. Med Phys 1999;26:2223 30.
[4] Klingenbeck-Regn K, Schaller S, Flohr T, et al. Subsecond multi-slice computed tomography: basics and
applications. Eur J Radiol 1999;31:110 24.
[5] Ohnesorge B, Flohr T, Becker C, et al. Cardiac imaging
by means of electrocardiographically gated multisection
spiral CT: initial experience. Radiology 2000;217:
564 71.
[6] Becker CR, Ohnesorge BM, Schoepf UJ, Reiser MF.
Current development of cardiac imaging with multidetector-row CT. Eur J Radiol 2000;36:97 103.
[7] Schoepf UJ, Becker CR, Obuchowski NA, et al. Multislice computed tomography as a screening tool for colon
cancer, lung cancer and coronary artery disease. Eur
Radiol 2001;11:1975 85.
[8] Flohr T, Bruder H, Stierstorfer K, et al. New technical
developments in multislice CT, part 2: sub-millimeter
16- slice scanning and increased gantry rotation speed
for cardiac imaging. Rofo Fortschr Geb Rontgenstr
Neuen Bildgeb Verfahr 2002;174:1022 7.
[9] Kopp AF, Kuttner A, Heuschmid M, et al. Multidetector-row CT cardiac imaging with 4 and 16 slices for
coronary CTA and imaging of atherosclerotic plaques.
Eur Radiol 2002;12(suppl 2):S17 24.
[10] Hu H, He HD, Foley WD, Fox SH. Four multidetectorrow helical CT: image quality and volume coverage
speed. Radiology 2000;215:55 62.
647
[11] Schoepf UJ, Becker CR, Bruening RD, et al. Electrocardiographically gated thin-section CT of the lung.
Radiology 1999;212:649 54.
[12] Becker CR, Jakobs TF, Aydemir S, et al. Helical and
single-slice conventional CT versus electron beam CT
for the quantification of coronary artery calcification.
AJR Am J Roentgenol 1999;174:543 7.
[13] Becker CR, Knez A, Jakobs TF, et al. Detection and
quantification of coronary artery calcification with electron-beam and conventional CT. Eur Radiol 1999;9:
620 4.
[14] Hong C, Becker CR, Schoepf UJ, et al. Coronary artery calcium: absolute quantification in nonenhanced
and contrast-enhanced multi-detector row CT studies.
Radiology 2002;223:474 80.
[15] Fleischmann D, Rubin GD, Bankier AA, Hittmair K.
Improved uniformity of aortic enhancement with customized contrast medium injection protocols at CT
angiography. Radiology 2000;214:363 71.
[16] Bae KT, Tran HQ, Heiken JP. Multiphasic injection
method for uniform prolonged vascular enhancement
at CT angiography: pharmacokinetic analysis and
experimental porcine model. Radiology 2000;216:
872 80.
[17] Hittmair K, Fleischmann D. Accuracy of predicting and
controlling time-dependent aortic enhancement from a
test bolus injection. J Comput Assist Tomogr 2001;25:
287 94.
[18] Napel S, Marks MP, Rubin GD, et al. CT angiography
with spiral CT and maximum intensity projection.
Radiology 1992;185:607 10.
[19] Wexler L, Brundage B, Crouse J, et al. Coronary artery
calcification: pathophysiology, epidemiology, imaging
methods, and clinical implications. A statement for
health professionals from the American Heart Association. Writing Group. Circulation 1996;94:1175 92.
[20] Shemesh J, Tenenbaum A, Fisman EZ, et al. Absence
of coronary calcification on double-helical CT scans:
predictor of angiographically normal coronary arteries
in elderly women? Radiology 1996;199:665 8.
[21] Laudon DA, Vukov LF, Breen JF, et al. Use of electron-beam computed tomography in the evaluation of
chest pain patients in the emergency department. Ann
Emerg Med 1999;33:15 21.
[22] Georgiou D, Budoff MJ, Kaufer E, et al. Screening
patients with chest pain in the emergency department
using electron beam tomography: a follow-up study.
[23] Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary artery calcium using ultrafast
computed tomography. J Am Coll Cardiol 1990;15:
827 32.
[24] Detrano R, Hsiai T, Wang S, et al. Prognostic value of
coronary calcification and angiographic stenoses in patients undergoing coronary angiography [see comments]. J Am Coll Cardiol 1996;27:285 90.
[25] Schmermund A, Bailey KR, Rumberger JA, et al. An
algorithm for noninvasive identification of angiographic
three-vessel and/or left main coronary artery disease in
648
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]

symptomatic patients on the basis of cardiac risk and
electron-beam computed tomographic calcium scores.
Arad Y, Spadaro LA, Goodman K, et al. Predictive
value of electron beam computed tomography of the
coronary arteries: 19-month follow-up of 1173 asymptomatic subjects. Circulation 1996;93:1951 3.
OMalley PG, Taylor AJ, Jackson JL, et al. Prognostic
value of coronary electron-beam computed tomography for coronary heart disease events in asymptomatic
populations. Am J Cardiol 2000;85:945 8.
ORourke RA, Brundage BH, Froelicher VF, et al.
American College of Cardiology/American Heart
Association Expert Consensus document on electronbeam computed tomography for the diagnosis and prognosis of coronary artery disease. Circulation 2000;102:
126 40.
Carr JJ, Crouse III JR, Goff Jr. DC, et al. Evaluation of
subsecond gated helical CT for quantification of coronary artery calcium and comparison with electron
beam CT. AJR Am J Roentgenol 2000;174:915 21.
Goldin JG, Yoon HC, Greaser III LE, et al. Spiral
versus electron-beam CT for coronary artery calcium
scoring. Radiology 2001;221:213 21.
Shemesh J, Apter S, Rozenman J, et al. Calcification of
coronary arteries: detection and quantification with
double-helix CT. Radiology 1995;197:779 83.
Becker CR, Kleffel T, Crispin A, et al. Coronary artery
calcium measurement: agreement of multirow detector
and electron beam CT. AJR Am J Roentgenol 2001;176:
1295 8.
Knez A, Becker C, Becker A, et al. Determination of
coronary calcium with multi-slice spiral computed tomography: a comparative study with electron-beam
CT. Int J Cardiovasc Imaging 2002;18:295 303.
Kopp AF, Ohnesorge B, Becker C, et al. Reproducibility and accuracy of coronary calcium measurements
with multi-detector row versus electron-beam CT. Radiology 2002;225:113 9.
Schmermund A, Erbel R, Silber S. Age and gender
distribution of coronary artery calcium measured by
four-slice computed tomography in 2,030 persons with
no symptoms of coronary artery disease. Am J Cardiol
2002;90:168 73.
Achenbach S, Ropers D, Mohlenkamp S, et al. Variability of repeated coronary artery calcium measurements
by electron beam tomography. Am J Cardiol 2001;87:
210 3.
Callister TQ, Raggi P, Cooil B, et al. Effect of HMGCoA reductase inhibitors on coronary artery disease
as assessed by electron-beam computed tomography.
N Engl J Med 1998;339:1972 8.
Shemesh J, Apter S, Stroh CI, et al. Tracking coronary
calcification by using dual-section spiral CT: a 3-year
follow-up. Radiology 2000;217:461 5.
Ohnesorge B, Flohr T, Fischbach R, et al. Reproducibility of coronary calcium quantification in repeat examinations with retrospectively ECG-gated multisection
spiral CT. Eur Radiol 2002;12:1532 40.
[40] Jakobs TF, Becker CR, Ohnesorge B, et al. Multislice

helical CT of the heart with retrospective ECG gating:
reduction of radiation exposure by ECG-controlled
tube current modulation. Eur Radiol 2002;12:1081 6.
[41] Hong C, Bae KT, Pilgram TK, et al. Coronary artery
calcium measurement with multi-detector row CT: in
vitro assessment of effect of radiation dose. Radiology
2002;225:901 6.
[42] Callister TQ, Cooil B, Raya SP, et al. Coronary artery
disease: improved reproducibility of calcium scoring
with an electron-beam CT volumetric method. Radiology 1998;208:807 14.
[43] Ulzheimer S, Kalender WA. Assessment of calcium
scoring performance in cardiac computed tomography.
Eur Radiol 2003;13:484 97.
[44] Becker C, Schoepf UJ, Reiser MF. Coronary artery calcium scoring: medicine and politics. Eur Radiol 2003;
13:455 7.
[45] Lee VS, Resnick D, Bundy JM, et al. Cardiac function:
MR evaluation in one breath hold with real-time true
fast imaging with steady-state precession. Radiology
2002;222:835 42.
[46] Juergens KU, Grude M, Fallenberg EM, et al. Using
ECG-gated multidetector CT to evaluate global left
ventricular myocardial function in patients with coronary artery disease. AJR Am J Roentgenol 2002;179:
1545 50.
[47] Ropers D, Ulzheimer S, Wenkel E, et al. Investigation
of aortocoronary artery bypass grafts by multislice spiral computed tomography with electrocardiographicgated image reconstruction. Am J Cardiol 2001;88:
792 5.
[48] Pump H, Moehlenkamp S, Sehnert C, et al. Electronbeam CT in the noninvasive assessment of coronary
stent patency. Acad Radiol 1998;5:858 62.
[49] Becker CR, Schoepf UJ, Reiser MF. Methods for quantification of coronary artery calcifications with electron
beam and conventional CT and pushing the spiral CT
envelope: new cardiac applications. Int J Cardiovasc
Imaging 2001;17:203 11.
[50] Nieman K, Oudkerk M, Rensing BJ, et al. Coronary
angiography with multi-slice computed tomography.
Lancet 2001;357:599 603.
[51] Knez A, Becker CR, Leber A, et al. Usefulness of
multislice spiral computed tomography angiography
for determination of coronary artery stenoses. Am J
Cardiol 2001;88:1191 4.
[52] Achenbach S, Giesler T, Ropers D, et al. Detection of
coronary artery stenoses by contrast-enhanced, retrospectively electrocardiographically-gated, multislice
spiral computed tomography. Circulation 2001;103:
2535 8.
[53] Becker CR, Knez A, Leber A, et al. Detection of coronary artery stenoses with multislice helical CT angiography. J Comput Assist Tomogr 2002;26:750 5.
[54] Kopp AF, Schroeder S, Kuettner A, et al. Non-invasive
coronary angiography with high resolution multidetector- row computed tomography: results in 102 patients.
Eur Heart J 2002;23:1714 25.

[55] Nieman K, Rensing BJ, van Geuns RJ, et al. Usefulness
of multislice computed tomography for detecting obstructive coronary artery disease. Am J Cardiol 2002;
89:913 8.
[56] Nieman K, Cademartiri F, Lemos PA, et al. Reliable
noninvasive coronary angiography with fast submillimeter multislice spiral computed tomography. Circulation 2002;106:2051 4.
[57] Hong C, Becker CR, Huber A, et al. ECG-gated reconstructed multi-detector row CT coronary angiography: effect of varying trigger delay on image quality.
Radiology 2001;220:712 7.
[58] Kopp AF, Schroeder S, Kuettner A, et al. Coronary
arteries: retrospectively ECG-gated multi-detector row
CT angiography with selective optimization of the
image reconstruction window. Radiology 2001;221:
683 8.
[59] Herzog C, Abolmaali N, Balzer JO, et al. Heart-rateadapted image reconstruction in multidetector-row
cardiac CT: influence of physiological and technical
prerequisite on image quality. Eur Radiol 2002;12:
2670 8.
[60] Nieman K, Rensing BJ, van Geuns RJ, et al. Non-invasive coronary angiography with multislice spiral computed tomography: impact of heart rate. Heart 2002;88:
470 4.
[61] Schroeder S, Kopp AF, Kuettner A, et al. Influence of
heart rate on vessel visibility in noninvasive coronary
angiography using new multislice computed tomography: experience in 94 patients. Clin Imaging 2002;26:
106 11.
[62] Giesler T, Baum U, Ropers D, et al. Noninvasive visualization of coronary arteries using contrast-enhanced
multidetector CT: influence of heart rate on image
quality and stenosis detection. AJR Am J Roentgenol
2002;179:911 6.
[63] Becker CR, Knez A, Ohnesorge B, et al. Imaging of noncalcified coronary plaques using helical CT with retrospective ECG gating. AJR Am J Roentgenol 2000;175:
423 4.
[64] Schroeder S, Kopp AF, Baumbach A, et al. Noninvasive detection of coronary lesions by multislice computed tomography: results of the New Age pilot trial.
Catheter Cardiovasc Interv 2001;53:352 8.
[65] Kopp AF, Schroeder S, Baumbach A, et al. Non-invasive characterisation of coronary lesion morphology and
composition by multislice CT: first results in comparison with intracoronary ultrasound. Eur Radiol 2001;11:
1607 11.
[66] Schroeder S, Kopp AF, Baumbach A, et al. Non-invasive characterisation of coronary lesion morphology by
[67]
[68]
[69]
[70]
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]
649
multi-slice computed tomography: a promising new

technology for risk stratification of patients with coronary artery disease. Heart 2001;85:576 8.
Schroeder S, Kopp AF, Baumbach A, et al. Noninvasive detection and evaluation of atherosclerotic coronary plaques with multislice computed tomography.
Botnar RM, Stuber M, Kissinger KV, et al. Noninvasive
coronary vessel wall and plaque imaging with magnetic
resonance imaging. Circulation 2000;102:2582 7.
Fayad ZA, Fuster V, Fallon JT, et al. Noninvasive in
vivo human coronary artery lumen and wall imaging
using black-blood magnetic resonance imaging. Circulation 2000;102:506 10.
Worthley SG, Helft G, Fuster V, et al. Serial in vivo
MRI documents arterial remodeling in experimental
atherosclerosis. Circulation 2000;101:586 9.
Fayad ZA, Fuster V, Nikolaou K, Becker C. Computed
tomography and magnetic resonance imaging for noninvasive coronary angiography and plaque imaging:
current and potential future concepts. Circulation
2002;106:2026 34.
Stary HC, Chandler AB, Dinsmore RE, et al. A definition of advanced types of atherosclerotic lesions and
a histological classification of atherosclerosis: a report
from the Committee on Vascular Lesions of the Council on Arteriosclerosis. American Heart Association.
Circulation 1995;92:1355 74.
Burke AP, Weber DK, Kolodgie FD, et al. Pathophysiology of calcium deposition in coronary arteries. Herz
2001;26:239 44.
Huang H, Virmani R, Younis H, et al. The impact of
calcification on the biomechanical stability of atherosclerotic plaques. Circulation 2001;103:1051 6.
Schmermund A, Schwartz RS, Adamzik M, et al. Coronary atherosclerosis in unheralded sudden coronary
death under age 50: histo-pathologic comparison with
healthy subjects dying out of hospital. Atherosclerosis 2001;155:499 508.
Virmani R, Kolodgie FD, Burke AP, et al. Lessons from
sudden coronary death: a comprehensive morphological
classification scheme for atherosclerotic lesions. Arterioscler Thromb Vasc Biol 2000;20:1262 75.
Virmani R, Burke AP, Farb A, Kolodgie FD. Pathology
of the unstable plaque. Prog Cardiovasc Dis 2002;44:
349 56.
Beckman JA, Ganz J, Creager MA, et al. Relationship of
clinical presentation and calcification of culprit coronary
artery stenoses. Arterioscler Thromb Vasc Biol 2001;
21:1618 22.
Radiol Clin N Am 42 (2004) 651 673
MR imaging in ischemic heart disease

Lawrence G. Dembo, MBBSa,b,*, Roger Y. Shifrin, MDc,
Steven D. Wolff, MD, PhDa,b,d
a
Advanced Cardiovascular Imaging, 62 East 88th Street, Lower Level, New York, NY 10128, USA
b
Cardiovascular Research Foundation, 55 East 59th Street, New York, 10022, NY, USA
c
Radiology Associates Imaging, Halifax Medical Center, 303 North Clyde Morris Boulevard,
Daytona Beach, FL 32114-2002, USA
d
Department of Radiology, Lenox Hill Hospital, 100 East 77th Street, New York, NY 10021, USA
Cardiovascular MR (CMR) imaging has become a

practical, clinically useful tool for the noninvasive
evaluation of ischemic heart disease. It allows for a
more comprehensive assessment of the heart than any
other single imaging modality. For example, CMR
can assess cardiac size and structure, global and
regional systolic function, myocardial perfusion and
viability, valvular morphology, flow, and coronary
artery anatomy. CMR offers high spatial and contrast
resolution (Fig. 1). One can acquire data in any
imaging plane with a large, unrestricted field of view.
Competing noninvasive technologies are each limited
by some combination of a requirement for iodinated
contrast, ionizing radiation, blurring from respiratory
motion, or the inability to image the entire heart
because of limited acoustic windows. This article
focuses on the expanding role MR imaging is playing
in the assessment of ischemic heart disease.
Functional evaluation
Cardiac size and function provide important prognostic information in ischemic heart disease and a
variety of acute and chronic cardiac diseases [1 3].
In clinical practice, ventricular volumes and global
function are usually subjectively assessed using two-
* Corresponding author. Advanced Cardiovascular

Imaging, 62 East 88th Street, Lower Level, New York,
NY 10128.
E-mail address: ldembo@mrict.com (L.G. Dembo).
dimensional echocardiography; however, both the

interobserver and intraobserver variation of CMR
are significantly lower than echocardiography [4].
Assessment of systolic ventricular function involves the evaluation of global and regional myocardial contraction. Regional myocardial function at rest
and following inotropic stimulation are better prognostic determinants of functional recovery following
revascularization than global measures of left ventricular (LV) function. Each segment must be visualized, and its motion characterized. A standardized
model for tomographic imaging of the heart involves
17 segments (Fig. 2) [5].
Because of the excellent contrast between the
endocardial surface and the blood pool, cardiac MR
imaging enables accurate and reproducible delineation
of the LV cavity [6 9] and is considered by many to
be the current standard of reference for the noninvasive assessment of cardiac function. Standard twodimensional transthoracic echocardiography requires
geometric assumption for volumetric measurements
(eg, Simpsons rule or the prolate ellipsoid method for
calculating left atrial volume). Although the normal
heart often has a high degree of symmetry, cardiac
pathology often distorts these patterns, making geometric assumptions less accurate. In contradistinction,
MR imaging data are three dimensional, which allows
direct measurement without the requirement of geometric assumptions. There is substantially less interstudy variation with CMR than with transthoracic
echocardiography (Table 1) when evaluating parameters of ventricular function [4,10,11]. Table 1 shows
the interstudy standard deviation percent for echocar-
doi:10.1016/j.rcl.2004.03.007
652
L.G. Dembo et al / Radiol Clin N Am 42 (2004) 651673
Fig. 1. Large, unrestricted field-of-view. Oblique coronal view prescribed parallel to the aorta (along the white line as shown
in the inset). This view is in an imaging plane not attainable by echocardiography. The heart, with a jet of aortic insufficiency,
is clearly visualized as are abdominal viscera and both lung fields. In this case, there is a lung mass in the right lower lobe
diography and CMR examination in both normal

hearts and abnormal hearts. There is little difference
in standard deviation percent when comparing the two
CMR groups. Transthoracic echocardiography has a
significantly higher standard deviation percent than
CMR, probably reflecting its lower image quality and
the geometric assumptions required to generate volumetric sets from two-dimensional data.
MR imaging acquisitions based on steady-state
free-precession (SSFP) are less dependent on inflow
enhancement than previous pulse sequences, and
have become the sequences of choice for evaluating
ventricular function, especially in patients with poor
systolic function and slow flow. Contrast is primarily
determined by the (M[T2/T1]) of the tissues, which is
significantly higher for blood than for myocardium.
This allows for high contrast (Fig. 3) between the
blood pool and myocardium throughout the cardiac
cycle, and consequently highly reliable and accurate
assessment of global and regional myocardial function [12 14]. SSFP images show substantial improvements in image quality, with less intraobserver
and interobserver variability compared with older
pulse sequences based on T1-weighted segmented
fast spoiled gradient echo [15,16]. Because of reliance on flow-related enhancement for contrast, these
older sequences are more susceptible to signal loss
caused by in-plane saturation and turbulent flow
[13,16,17]. Both segmented fast spoiled gradient
echo and SSFP techniques provide qualitative assessment of myocardial motion with adequate (eg, < 80
millisecond) temporal resolution. SSFP images may
also be acquired in real time to provide reproducible,
accurate evaluation of ventricular function and mass,
without the need for normal sinus rhythm, breathholding, or cardiac gating [18].
Left ventricular function can be assessed qualitatively and quantitatively. Qualitative assessment
involves a subjective evaluation of function both
globally and regionally. Abnormal segments are characterized as hypokinetic (less than 40% systolic
thickening); akinetic (less than 10% systolic thickening); dyskinetic (paradoxical systolic motion with
thinning); or aneurysmal.
Several MR imaging methods are available quantitatively to assess myocardial motion. Normal LV
myocardium demonstrates systolic wall thickening
and radial shortening, with measures of regional
653
Fig. 2. Standard segmental nomenclature. Series of short axis diagrams indicating how the left ventricle can be divided into
17 segments. The heart is divided into three sections of equal longitudinal length: base, mid-ventricle, and apex. The base and
midventricle are each divided into six equal segments, the apex into four. The tip of the apex is designated as segment 17. This
17-segment model is generally accepted by many professional cardiac imaging societies.
myocardial function including systolic wall thickening, wall motion, and myocardial strain. Tagging
methods apply a parallel grid, two-dimensional array,
or radial pattern of radiofrequency saturation bands
to the myocardium before excitation and readout
Table 1
Sample size: MRI versus transthoracic echocardiography
Standard deviation percent
Ejection fraction
End-systolic
volume
End-diastolic
volume
Myocardial mass
CMR
normal
CMR
abnormal
Echocardiography
2.4
4.7
2.5
6.5
6.6
15.8
3.5
7.4
23.8
6.4
6.4
36.4
Abbreviation: CMR, cardiovascular magnetic resonance.

(Adapted from Bellenger NGL, Davies C, et al. Reduction in
sample size for studies of remodeling in heart failure by the
use of cardiovascular magnetic resonance. J Cardiovasc
Magn Reson 2000;2:271 8; with permission.)
[19]. The applied tags can be tracked throughout

the cardiac cycle to evaluate qualitatively and quantitatively myocardial motion [20 29]. Either tagging
methods or cine phase-contrast techniques can be
used to measure local myocardial tissue deformation
(strain) throughout the cardiac cycle. Although a
detailed analysis of the full strain tensor may be
useful for precise understanding of myocardial physiology, it is mathematically and computationally
intensive and is not practical for routine clinical
practice. Semiautomated techniques, such as harmonic-phase MR imaging [30], strain encoding [31],
velocity encoding [32,33], and displacement encoding with stimulated echoes [34 36], are being developed to integrate the quantitative strain measures into
a clinically useful tool [37].
Significant coronary stenosis

There are two common approaches to the noninvasive assessment of significant epicardial coronary
654
Fig. 3. CMR steady-state free precession images. (A) Four-chamber, (B) three-chamber, (C) short-axis views, and (D) oblique
coronal. All show excellent contrast between the blood pool and myocardium.
stenosis: stress perfusion testing and stress function testing.

Perfusion
Myocardial perfusion is used as a test for evaluating the adequacy of blood flow to the myocardium.
Because of its noninvasive nature, it is often used as a
surrogate to detect coronary artery disease. Perfusion
analysis, which measures the adequacy of blood flow,
is fundamentally different than assessing coronary
lumenography as descried by coronary angiography.
For this reason, correlation between coronary angiography and perfusion is not exact. For example,
although an epicardial coronary artery may be occluded, perfusion may be normal if supplied by
collaterals from a different vessel. Alternatively, an
eccentric stenosis can be missed by coronary angiography or diffuse narrowing of the lumen may be
recognized but not thought to be significantly ste-
nosed, yet both may exhibit perfusion defects. The

clinical evaluation of myocardial perfusion is most
commonly performed by single-photon emission CT
(SPECT) and less commonly by positron emission
tomography (PET) or contrast echocardiography.
Each of these techniques suffers from some combination of requirement for radioactive tracer, long examination times, poor acoustic windows, and limited
spatial resolution. Numerous studies have validated
the accuracy of the CMR technique as compared with
radiolabeled microspheres, SPECT, PET, and coronary angiography [38 48].
Physiologic principles of perfusion imaging
At rest, the blood flow to myocardium distal to a
stenotic epicardial coronary artery lesion can have
normal perfusion because of selective arteriolar vasodilation. This vasodilatory process is termed autoregulation (Fig. 4). The extent of the maximal
increase in blood flow defines the myocardial perfu-
Fig. 4. Autoregulation, a schematic. (A) Rest: arteriolar segments distal to a stenotic parent vessel dilate by autoregulatory mechanisms to maintaining distal flow. The normal
arteriolar segment retains contractile reserve and has the
capacity to dilate in response to increased oxygen demand or
pharmacologic vasodilation. (B) Stress: the normal vessel
dilates allowing increased coronary flow. Without perfusion
reserve the vessel distal to a stenotic lesion remains the same
size resulting in relatively delayed and diminished coronary
flow into that territory.
sion reserve [49]. At rest, the process of autoregulation may result in equal perfusion of myocardial
segments supplied by a stenotic coronary artery and
segments supplied by a normal coronary artery. In
these regions of myocardium distal to a stenotic lesion
there is diminished capacity to increase blood flow in
the face of increased oxygen demand or generalized
coronary arteriolar dilation and consequent loss of
myocardial perfusion reserve. A segment distal to a
stenosis may then become hypoperfused relative to
the normal segment during periods of increased coronary blood flow [50,51]. This is the basis for stress
perfusion testing. Stress may be induced by increasing
demand by exercise or pharmacologically using inotropes, such as dobutamine. Pharmacologic vasodilators, such as adenosine or dipyridamole, increase
resting coronary flow four to eight times in regions
of normal perfusion [52,53]. These vasodilators do not
induce ischemia by increasing myocardial oxygen
demand and are probably safer in an outpatient
setting. Also, particularly adenosine has a rapid onset and cessation of action allowing for quick and
safe testing.
Vasodilator stress perfusion
Regional myocardial blood flow may be assessed
by dynamic MR imaging during the first pass of an
extracellular gadolinium chelate contrast agent. It is
important to obtain complete coverage of the heart
with high temporal resolution adequately to track the
passage of contrast through the myocardium. Compared with nuclear perfusion studies, CMR has the
advantage of being able to evaluate perfusion across
the transmural extent of the myocardium, allowing
differentiation of the subepicardium and subendocar-
655
dium. Currently, clinically used sequences tend to be

cardiac gated and T1-weighted, such as fast gradient
echo (FGE) and fast gradient echo-echo train imaging
(FGRET). Newer perfusion pulse sequences with
improved contrast based on SSFP are becoming
available [54]. The number of slices to obtain good
coverage of the heart form base to apex is related to
heart rate, cardiac size, hardware, and the specific
sequence used. For a heart rate of about 90 beats per
minute, current clinical protocols acquire up to five
short-axis slices per heartbeat or 10 short-axis slices
every two heartbeats. First-pass imaging is routinely
completed in less than a minute. The high temporal
resolution does not predicate breathholding; however,
patients often breathhold to improve image registration over the imaging time course.
Following bolus infusion of contrast, there is
sequential progression of contrast through the right
ventricle, lungs, left ventricle, and myocardium. This
appears as enhancing bright signal within the cardiac
chambers, myocardium. In segments of myocardium
supplied by a stenotic coronary artery, there is characteristically delayed and diminished myocardial enhancement in the presence of pharmacologic stress
(Fig. 5), but normal enhancement at rest. Myocardial
infarction shows characteristic delayed and diminished enhancement at rest and with stress. Within the
perfusion defect there is a gradient from subepicardium toward the darker subendocardium, which aids
in differentiation from artifact.
There has been considerable interest in MR imaging first-pass perfusion techniques in the setting of
acute ischemia [55]. First-pass MR imaging accurately can assess regions of nonviable microvascular
obstruction (no reflow) following acute myocardial
infarction. The size of the zone of microvascular
obstruction increases over the first 2 days after
reperfusion, and is related to the infarct size and
the duration of coronary artery obstruction before
reperfusion [56 59]. The extent of microvascular
obstruction, as determined by MR imaging, predicts
Fig. 5. Delayed and diminished perfusion. Ischemic myocardial segments exhibit delayed and diminished enhancement following bolus contrast administration relative to
normally perfused myocardial segments.
656
long-term prognosis [60], return to function [60], and

occurrence of cardiovascular events over a 2-year
follow-up period [61]. Using an SSFP adenosine
stress perfusion protocol, Chiu et al [62] showed that,
in acute coronary syndromes without elevation of
creatine kinase (ie, unstable angina), MR imaging
was 92% sensitive and 92% specific in detecting a
greater than 50% stenosis as defined by cardiac
catheterization. Similar studies have evaluated different combinations of first-pass perfusion MR imaging,
functional MR imaging, and myocardial delayed
enhancement (MDE) imaging to evaluate patients
with acute coronary syndromes [63 65].
Analysis
Qualitative analysis. First-pass perfusion imaging
can be analyzed on a qualitative or quantitative basis.
Qualitative analysis is straightforward (Fig. 6), and
can be completed in a relatively short time. The
accuracy of qualitative first-pass perfusion imaging
has been demonstrated in several recent trials. Using
dipyridamole stress MR imaging, Ishida et al [66]
studied 104 patients, including 69 who also underwent SPECT, and compared the results with coronary
angiography. MR imaging was significantly better
than SPECT for detection of a 70% stenosis (area
under the curve = 0.9 for MR imaging, 0.73 for
SPECT; P < .001). The sensitivity of MR imaging for
the detection of single-, double-, and triple-vessel disease was 85%, 96%, and 100%, respectively. The overall sensitivity and specificity to detect at least one
coronary artery with a significant stenosis was 90%

and 85%, respectively. Similar results were demonstrated in a multicenter study of 99 patients using MR
imaging with adenosine compared with quantitative
coronary angiography. MR imaging perfusion studies
were evaluated subjectively by four blinded
reviewers with average area under the receiver operating characteristic curve (ROC) curve of 0.90 when
compared with quantitative coronary angiography.
In this group, the sensitivity, specificity, and accuracy
to detect a 70% stenosis were 93%, 75%, and 85%
respectively (Wolff et al, submitted for publication).
Quantitative analysis. Although the concept of true
quantitative analysis is attractive, there are multiple
limitations including assumptions of an ideal impulse
arterial input function and linearity of signal with respect to gadolinium concentration, cardiac and respiratory motion, signal falloff because of the use of
surface coils, susceptibility artifacts, and artifacts
related to parallel imaging techniques. Because of
these difficulties, investigators have focused on the
analysis of myocardial perfusion reserve or myocardial perfusion reserve index, which is the ratio of
the rates of change in signal intensity during stress
compared with rest.
Quantitative techniques rely on the accurate assessment of the signal intensity measurements during
the early phases of the first-pass of contrast. Because
peak signal intensity is not linearly related to gadolinium concentration [67], the rate of change of signal
Fig. 6. Perfusion defect. Selected short axis views obtained during pharmacologic stress (top row) and during resting conditions
(bottom row). The apex of the heart is to the viewers left with successive images to the right more toward the base. These images
illustrate perfusion defects as evidenced by diminished enhancement in the subendocardial zone of the lateral wall and extending
into the anterior wall and inferior wall toward the base of the heart.
657
intensity is more reliable for the detection of coronary

artery stenosis [68]. Standardization is performed by
correcting for precontrast signal intensity and using
the ratio of the corrected upslope of signal intensity
in the region of interest to the corrected upslope of
LV signal intensity. First-pass perfusion imaging with
MR imaging can be performed at rest and during
stress with comparison of the standardized upslopes
in the region of interest to calculate the myocardial
perfusion reserve index. The exact myocardial perfusion reserve index is very difficult to calculate;
however, there are simplified models available, for
example using only a few time points in the upslope
curve, which provide close approximations.
A number of recent studies have used the myocardial perfusion reserve index to evaluate myocardial
perfusion. In a prospective study of 34 subjects, dipyridamole stress MR imaging had a sensitivity, specificity, and diagnostic accuracy of 90%, 83%, and
87%, respectively, for detecting significant stenosis
(> 75% at cardiac catheterization) [69]. In another
study, 84 patients referred for a primary diagnostic
catheterization underwent adenosine stress MR imaging, which demonstrated a sensitivity, specificity, and
accuracy of 88%, 90%, and 89%, respectively [70].
Similar results have been reported in smaller studies
[71 74].
In a comparison between stress MR imaging
analyzed semiquantitatively, PET, and cardiac catheterization, MR imaging demonstrated a sensitivity
and specificity of 91% and 94%, respectively, for
the detection of coronary artery disease as defined by
PET and a sensitivity and specificity of 87% and
85%, respectively, in comparison with quantitative
coronary angiography [75]. In addition, the extent of
disease, as defined by the number of pathologic
sectors per patient, was highly correlated (slope =
0.94, P < .0001) when MR imaging and PET were
compared. A smaller study comparing coronary
flow reserve measurements by adenosine stress MR
imaging and PET also showed a close relationship
between MR imaging upslope index and PET estimation of flow reserve [76].
monly) two-dimensional transthoracic echocardiography. In normal myocardial segments inotropic

stimulation increases regional function and causes
an increased myocardial oxygen demand with a corresponding increase in blood flow. In segments distal
to a stenotic coronary artery, there is limited flow
reserve. A myocardial segment supplied by a stenotic
artery may demonstrate normal wall motion at rest,
but develop wall motion abnormalities following
inotropic stimulation. Similar in concept is the biphasic response, defined by an initial improvement in
myocardial contraction with a low level of stress,
followed by a decline in myocardial function (once
perfusion reserve has been depleted) at higher levels
of stress.
Hundley et al [79] using dobutamine stress cine
MR imaging studied 153 patients who had poor
acoustic windows at echocardiography. The sensitivity and specificity for detection of a significant
coronary artery stenosis (> 50% luminal narrowing)
as compared with quantitative coronary angiography
were both 83%. Significantly, a negative MR imaging
predicted a cardiovascular occurrence-free survival at
500 days in 97% of patients studied. Following these
patients and adding to the cohort, these investigators
have recently shown that in 279 patients, a positive
dobutamine stress MR imaging predicted a greater
than threefold increase in the likelihood of myocardial infarction or death [80]. In a study of 208
patients, Nagel et al [81] compared dobutamine stress
MR imaging with dobutamine stress echocardiography and demonstrated that the sensitivity to detect
significant coronary disease increased from 74% to
86% and specificity increased from 70% to 86% with
stress MR imaging. They conclude that dobutamine
stress MR imaging yields significantly higher diagnostic accuracy than dobutamine stress echo, caused
mostly by the higher likelihood of visualizing all of
the myocardial segments in each patient undergoing
the test.
Stress function
The blood oxygen level dependent imaging technique exploits the paramagnetic properties of deoxyhemoglobin as an intrinsic contrast agent and does
not require exogenous contrast agents. In myocardial
segments supplied by a stenotic coronary artery, there
is maximal oxygen extraction from the capillary
bed as compared with normal myocardial segments
where there is submaximal oxygen extraction. The
fraction of paramagnetic deoxyhemoglobin in vessels
downstream of a stenosis is higher than in normal
Another method of noninvasively evaluating the

presence of coronary artery disease is by observing
myocardial segmental function at rest and in response
to increased demand for cardiac work. Potentially
ischemic myocardial segments may be identified
based on transient inducible wall motion abnormalities during pharmacologic (eg, dobutamine [77]) or
exercise [78] stress in conjunction with (most com-
Developing application: blood oxygen

level dependent imaging
658
tissue and can in principle be detected using a T2*weighted imaging sequence [82]. This blood oxygen
level dependent imaging technique is well established in animal [83,84] and human [85 88] cardiac
imaging. Wacker et al [89] measured the T2* at rest
and following dipyridamole stress. They showed
significant reductions of T2* in regions associated
with a stenotic epicardial artery (P < .01) with further
T2* reduction after vasodilator stress (P < .001). This
technique shows promise as a noninvasive measure
of myocardial perfusion without the requirement of
exogenous contrast.
Myocardial viability
Following ischemia, myocardium may be infarcted
or reversibly injured [90 92]. Reversible myocardial
dysfunction may be acute or chronic. Stunned
myocardium occurs following an acute ischemic
episode with early reperfusion. The muscle is dysfunctional but viable. Segmental dysfunction may
remain for up to 3 to 6 months after the ischemic
insult. Hibernating myocardium is viable but dysfunctional because of chronic ischemia [91]. Hibernating segments are likely to improve in function
following revascularization [93].
Hibernating, stunned, and infarcted myocardium
may all appear as regional wall motion abnormalities
of any degree. The ability to distinguish hibernating
or stunned myocardium from infarcted myocardium
is important as it may guide therapeutic intervention
and subsequent prognosis, as revascularization of
viable segments may improve regional and global
LV function [94 98] and subsequent long-term survival [2,99 103].
There are two methods used to assess myocardial
viability: contrast-enhanced MR imaging to recognize MDE and the identification of myocardial contractile reserve.
Delayed hyperenhancement
The second method of assessing myocardial viability is contrast-enhanced CMR and the evaluation
of MDE. Current MR imaging techniques for detecting myocardial viability rely on the extracellular
distribution of gadolinium chelates within the myocardium. Gadolinium distributes in the extracellular
space according to an open two-compartment pharmacokinetic model. In regions of increased extracellular space (eg, infarction), higher concentrations of
gadolinium accumulate with concomitant slower
clearance and higher signal on T1-weighted se-
quence. After acute myocardial insult, the extracellular volume in the infarcted region is increased about
fourfold [104,105]. With time, the extracellular volume begins to decrease but remains about double in
size compared with preinfarction because of chronic
inflammation and fibrosis [106]. In 2001 Simonetti et
al [107] showed an average of 485% increase in
signal intensity between infarcted and normal myocardium using a heavily T1-weighted inversion recovery sequence set to null normal myocardium. At
some delayed time following gadolinium administration (10 to 30 minutes), high spatial resolution images
are acquired during suspended respiration. Regions
with increased extracellular space (eg, infarction)
appear white, in contrast to the adjacent normal black
(nulled) myocardium. This pulse sequence, known as
MDE or delayed contrast-enhanced MR imaging,
has become the standard imaging sequence for the
assessment of myocardial infarction. This delayed
enhancement occurs in acute and chronic infarction,
which precludes the assessment of infarct age by this
technique. In some acute infarcts, however, microvascular obstruction may occur. In these cases the
core of the infarct remains dark, distinguishing it
from chronic infarction. Any entity that causes myocardial scar and subsequent increased extracellular
volume also shows similar contrast (eg, sarcoid myocarditis [108 114], acute myocarditis [115,116], and
hypertrophic cardiomyopathy [117 120]). Future
improvements include three-dimensional MDE
sequences, which offer increased signal-to-noise ratio
(SNR) and allow for the acquisition of the entire LV
length in a single breathhold.
The high spatial resolution of the MDE sequence
allows exquisite definition of the extent of injured
myocardium (Fig. 7). Numerous studies have demonstrated the excellent correlation between size, shape,
and volume of the hyperenhancement zone as compared with histopathology [58,121 126]. Gerber et al
[127] concluded that absence of delayed hyperenhancement had a sensitivity of 82% and an accuracy
of 74% in predicting recovery of myocardial function. It has also been shown that, for acute infarction,
the transmural extent of infarction as demonstrated by
MDE predicts long-term improvement in contractile
function [126,128,129]. Kim et al [130] showed a
striking relationship between the transmural extent of
hyperenhancement using MDE and the likelihood of
improvement of contractile function after revascularization. Globally, an increasing extent of delayed
hyperenhancement correlated with decreased improvement in the mean wall-motion score (P < .001)
and ejection fraction after revascularization (P <
.001). Regional functional improvement was also
659
Fig. 7. Microvascular obstruction MDE and anomalous coronary artery. (A) Short axis and (B) two-chamber MDE images of a
patient with a recent infarct (arrows). There is a transmural infarct involving the anterior wall in the mid-ventricle. The central
dark region of the infarct represents an area where extracellular contrast cannot diffuse and is characteristic of microvascular
obstruction. (C, D) Images from the same patient show an anomalous coronary artery (arrows) with malignant course between
the aorta and the pulmonary artery. Cardiac catheterization showed a long anomalous left main with occlusion of the first
diagonal branch of his left anterior descending coronary artery.
predicted by the transmural extent of delayed hyperenhancement in each segment ( P < .001). This is
independent of time since the ischemic insult, the
presence of wall motion abnormalities, or the history
of revascularization [128].
Importantly, MDE can define the presence, location, and transmural extent non Q wave myocardial
infarction [131]. Diagnosis of myocardial infarction
can be missed clinically if not recognized during the
relatively short period of cardiac enzyme elevation.
These small, usually subendocardial, infarcts are often
not associated with regional wall motion abnormalities and may not be diagnosed with conventional tests
of functional assessment. This is important because
non Q wave infarction is a sensitive marker for future
ischemic events [132]. Non Q wave infarcts have a
higher incidence [133], and have a mortality rate that
is equal to [134,135] or greater than [136] Q wave
infarction. The MDE technique allows detection of
even the small areas of infarction known to occur

after some percutaneous coronary interventions
[137 140]. This has important prognostic implications because even small creatine kinase rises post
percutaneous coronary interventions are associated
with poorer long-term outcomes [141,142].
Myocardial delayed enhancement: comparison with
other modalities
Single-photon emission CT
Nuclear myocardial perfusion imaging is the most
commonly ordered noninvasive test to assess myocardial viability. The spatial resolution of SPECT is
limited. Voxel volumes are approximately 50 times
the typical voxel size of MR imaging, and are on the
same order as average myocardial wall thickness.
Although a reduction in voxel size on the newer
SPECT cameras may decrease the voxel size differ-
660
ential, the effective resolution of MR imaging is

much higher than SPECT because the SPECT suffers
from a large amount of blur caused by respiratory
motion. Some studies suggest that SPECT misses
small infarcts probably because of its lower resolution [143 145]. An infarct smaller than one voxel in
size is volume averaged with adjacent normal myocardium. This limits the sensitivity of nuclear SPECT
imaging for detecting nontransmural and small transmural myocardial infarction. MDE has an in-plane
spatial resolution of approximately 1.5 mm, which
permits assessment of these small infarctions and
allows evaluation of the extent of infarction relative
to myocardial wall thickness.
Wagner et al [146] compared MDE and SPECT
and concluded MDE systematically detects subendocardial infarcts that are missed by SPECT
(Fig. 8). In that study of 91 patients, 47% of segments
with subendocardial infarction detected by MDE
were not detected by SPECT. Unless diagnosed in
the acute setting, these patients usually have no other
evidence of myocardial infarction and may not have
been diagnosed and treated for ischemic heart disease. Untreated survivors of myocardial infarction
have a mortality rate of about 14 times that of the
normal population [147]. With SPECT imaging
alone, patients with small myocardial infarcts could
have false-negative studies, not be offered secondary
prevention, and possibly suffer poorer clinical out-
comes. Kitagawa et al [148] showed similar results in

a study of 22 patients early after acute myocardial
infarction. In this study, the sensitivity (98% versus
90.3%, P < .01), specificity (75% versus 54.4%, P <
.05), and accuracy (92% versus 81.1%, P < .001) of
MDE in the prediction of viable myocardium were
significantly higher than those of resting thallium 201
SPECT, the results of which match closely with
published cumulative average sensitivity and specificity data [149].
Positron emission tomography
Metabolic imaging with 18F-fluorodeoxyglucose
(FDG) PET has, in the past, been regarded as the
standard of reference for the detection of myocardial
viability. Assessment of the extent of viable myocardium with PET correlates well with improved contractile performance after revascularization and both
short- and long-term prognosis [150 153]. Kuhl et al
[154] compared MDE with PET in patients with
ischemic heart disease and severe LV dysfunction.
They conclude that MDE allows assessment of myocardial viability with a high accuracy compared with
FDG PET. Klein et al [151] studied 31 patients with
severe heart failure and concluded that MR imaging
enhancement closely agrees with PET data; however,
MDE identifies scar tissue more frequently than PET
reflecting its higher spatial resolution.
Identification of contractile reserve: stress MR
imaging
Fig. 8. Transmural extent. Short-axis MDE image illustrating

the high-resolution images of CMR, which allows infarctreporting based on their transmural extent. Here, there is a
large transmural infarct (long arrow) involving the anterior
wall and extending into the adjacent subendocardium (small
arrows) of the anteroseptal and anterolateral walls. Also
shown is a smaller subendocardial infarction involving the
inferolateral wall (small arrow, thicker arrowhead).
Cardiovascular MR imaging may be used to

detect hibernating or stunned myocardium. Hibernating myocytes, by definition, can increase contractile
function following inotropic stimulation [155]. The
presence of a contractile reserve provides a method
for the noninvasive detection of reversible myocardial dysfunction. An ischemic, but viable myocardial
segment may be akinetic, hypokinetic, or dyskinetic
at rest, but because of the presence of viable myocytes may exhibit increased systolic wall thickening
and radial shortening following inotropic stimulation.
Sandstede et al [156] studied patients with akinetic or
dyskinetic myocardial segments by MR imaging at
rest and following dobutamine administration and
demonstrated a positive predictive value for viability
of 87% on a per segment basis, and 100% on a per
patient basis. Detection of viable myocardium is not
only important for surgical planning before revascularization, but is also an important prognostic
indicator following myocardial infarction. Early intervention to restore blood flow is the treatment of
choice in patients with acute myocardial infarction.
Assessment of viable but dysfunctional myocardium

(stunned myocardium) after revascularization is useful to evaluate the adequacy of revascularization
therapy. Baer et al [157] compared dobutamine MR
imaging with F18-FDG PET in 35 patients with a
history of myocardial infarction. In this study, the
sensitivity, specificity, and positive predictive value
of a dobutamine-induced contraction reserve for
detection of viable myocardium as defined by PET
were 81%, 95%, and 96%, respectively.
Thrombus
Thrombus may form in association with regions
of myocardial contractile dysfunction. Thrombus
within the cardiac chambers can be missed on transthoracic echocardiography because of inadequate
acoustic windows or poor conspicuity of thrombus
relative to normal myocardium. MDE (Fig. 9) and
SSFP imaging sequences are more sensitive than
transthoracic echocardiography for the detection of
intracardiac thrombus [158,159]. Furthermore, MR
imaging has the potential to differentiate subacute
thrombi, which tend not to enhance after contrast,
from late organizing thrombus, which tends to enhance after contrast [160]. This is a challenging task
for echocardiography and is clinically important
because subacute thrombus is more likely to embolize [159].
Analysis of coronary arteries

Although MR angiography is performed routinely
for non coronary artery applications, imaging of the
coronary arteries lumen is complicated by multiple
factors, most importantly cardiac motion, respiratory
661
motion, and fat suppression. The size of the vessel

being imaged (0.5 4 mm) is much smaller than the
vessel excursion during the cardiac and respiratory
cycles. This necessitates methods to compensate for
the complex displacement of the coronary arteries
because of bulk cardiac motion and respiratory motion [161].
Motion compensation
A variety of techniques have been used to minimize the effect of cardiac motion. These include
prospective cardiac triggering, cardiac navigator
sequences, and real-time MR imaging. Cardiac triggering requires accurate monitoring of the cardiac
cycle, which may be accomplished with ECG leads,
peripheral pulse, or navigator techniques [162,163].
Because of the displacement of the coronary
arteries throughout the respiratory cycle, imaging is
performed during suspended respiration or using free
breathing with data acquisition limited to a specific
respiratory phase (typically end-expiration). Breathhold techniques have the advantage of a shorter
image acquisition time, are limited by the patients
ability to hold their breath and diaphragmatic drift
[164], and subtle relative phase shifts caused by chest
wall motion [165]. The limited time a patient can
breathhold limits spatial resolution. If image time is
longer than a comfortable breathhold, then imaging
may be extended over multiple separate breathholds
because slice misregistration between adjacent
acquisitions may occur if there is any variation in
position within the imaging slice during serial breathholds [166,167].
Images that are acquired with the patient free
breathing are easier for patients; however, they take
longer to acquire and require software to acquire data
Fig. 9. MDE thrombus. (A) Three-chamber and (B) short axis apical MDE images show a large transmural apical infarction with
thrombus adherent to the endocardial apical surface. The MDE technique is the most sensitive CMR technique for detecting
intracardiac thrombus.
662
accurately during a certain phase of the respiratory

cycle. Navigator techniques may be used to limit data
acquisition to a short period of the respiratory cycle
during which respiratory motion is minimal. This can
involve excitation of a target tissue (eg, diaphragm,
LV, epicardial fat), which is then monitored throughout the respiratory cycle. Data acquisition is then
limited to a narrow, user-defined, acceptance window.
Free breathing allows the acquisition of higher spatial
resolution scans, higher SNR, and the ability to image
patients who are unable to hold their breath. Sophisticated navigator sequences are being developed,
such as hybrid techniques, which add an initial
breathhold, to a free breathing navigator sequence,
such that the center of k-space is acquired during
suspended respiration, followed by navigator-monitored free breathing [168]. Advances to the navigator
technique including motion-adapted gating [169
171], slice tracking, and sophisticated motion analysis, such as the diminishing variance algorithm
[172,173], should translate into increased diagnostic
accuracy. The main drawback of all of the navigator
techniques is the accuracy to acquire data at the
same point in the respiratory cycle over the duration
of the scan.
Technical factors
A variety of techniques have been used for coronary MR angiography including two- and threedimensional segmented k-space gradient echo [174],
multishot echo planar imaging (EPI) [175], and spiral
k-space acquisitions. Most current efforts use threedimensional techniques, which offer the advantage
of higher SNR and spatial resolution and greater potential for postprocessing using volume rendering,
maximum intensity projection, or curved planar reformation [176,177]. The goal of all of these techniques is to make the artery lumen brighter and
simultaneously suppress the surrounding fat.
Spiral k-space acquisition is attractive because of
its efficient use or gradient power, allowing for
reduced scan times [178]. Furthermore, the spiral
k-space acquisition offers greater sampling density
at the center of k-space and is relatively insensitive to
motion-induced phase errors because first gradient
moments are inherently zero [179]. Cartesian k-space
acquisitions with asymmetric sampling also reduce
scan time and are less sensitive to off-resonance
phase errors [180]. These innovative sequences will
likely further improve acquisition time, SNR, and
CNR [176,181]. Projection reconstruction techniques
may also be used for coronary MR angiography
providing higher spatial resolution in a shorter imaging time [182].

Parallel encoding techniques (sensitivity encodingarray spatial sensitivity coding [SENSE-ASSET],
simultaneous acquisition of spatial harmonics
[SMASH], and newer generalized methods [183
185]) allow the acquisition of fewer lines of k-space
exploiting the spatial information contained within the
different physical location of the individual elements
of a multicoil array [184 186]. Using these parallel
data acquisition schemes, images including coronary
artery images can be acquired more quickly but with
less SNR [18,187,188].
The introduction of new whole-body 3-T MR
imaging systems has generated interest in cardiac
MR imaging at ultrahigh field strength. In particular
the higher SNR afforded by 3-T systems should
translate into higher CNR and higher spatial resolution coronary MR angiography with improved branch
vessel visualization. The feasibility of coronary MR
angiography at 3 T in vivo has been demonstrated
[189]. Existing pulse sequences have to be adapted to
the higher field strength to avoid excessive radiofrequency deposition and to minimize the effects of
the shorter T2* at high field. Furthermore, because of
the magnetohydrodynamic effect, gating schemes
based on ECG amplitude alone are less effective
because T-wave swelling may approximate the amplitude of the QRS complex.
Contrast
The key to coronary MR imaging is obtaining
good contrast between bright signal in the lumen and
dark signal in the surrounding fat and muscle. Early
attempts at coronary MR angiography relied on flowrelated enhancement for vascular signal [190 193],
which is limited by such factors as in-plane saturation.. Newer techniques to increase the signal from
the coronary lumen take advantage of the long T2 of
blood. SSFP sequences, which have high contrast
(M[T2/T1]) between the blood pool and myocardium,
are more immune to flow-related artifacts and to
some extent mitigate the problem of in-plane saturation. These sequences have been validated and shown
to have higher SNR and CNR than conventional
gradient echo sequences [194,195] enabling clearer
vascular wall and lumen definition. The relatively
short T2 of myocardium relative to coronary artery
blood provides a method further to modify image
contrast by suppression of the myocardial signal in
surrounding tissue. A T2 preparation pulse before an
SSFP sequence, for example, has been shown further
to improve tissue contrast [196 198] by making the
surrounding muscle dark while increasing the signal

from the coronary lumen (Fig. 10).
Extracellular contrast agents have been used to
improve the contrast between coronary arteries and
the adjacent myocardium and epicardial fat. Zheng
et al [199] demonstrated a three to five times increase
in coronary artery signal following administration of
an extracellular gadolinium chelate. Because these
agents rapidly redistribute into the extravascular
space and out of the artery, there is a limited time
window for data acquisition with optimal contrast,
and consequently a lower achievable spatial resolution. Newer contrast agents with longer intravascular
half lives (blood pool contrast agents) are being
investigated for coronary MR angiography. Stuber
et al [200] demonstrated a 69% increase in CNR
between myocardium and blood relative to baseline
T2 preparatory images using an investigational intravascular contrast agent (MS-325/AngioMARK, EPIX
Medical, St. Louis, Missouri).
Coronary artery bypass grafts
It is known that the patency of coronary artery
bypass grafts decrease with time. Long-term studies
show occlusion rates of 12% for vein grafts at 2 weeks
663
postsurgery, 25% at 1 year, and about 60% at 10 years

[201]. Reported occlusion rates for arterial grafts are
less than vein grafts (about 5% at 1 year and 10% at
10 years) [202]. Early bypass graft occlusion is
predominantly caused by graft thrombosis rather
than neointimal hyperplasia or atherosclerotic plaque
[203,204]. For the year 1997, it was estimated that
about 300,000 coronary artery bypass grafts procedures were performed in the United States. Approximately 30,000 to 40,000 patients per year suffer
acute graft occlusion.
A rapid, noninvasive test, such as MR imaging,
which can reliably detect coronary artery bypass graft
occlusion is an attractive option when compared with
conventional invasive catheter angiography. As opposed to native coronary arteries, it can be less
challenging to visualize vein grafts by MR imaging
because they have a larger diameter and are relatively
motionless throughout the cardiac cycle. Sequences
for native coronary MR angiography have improved
significantly in the last few years [205,206] and have
become potentially useful diagnostic tools. MR angiography can detect coronary artery bypass graft
occlusion, quantify and map flow, and visualize the
graft lumen with submillimeter resolution [207 211].
Challenges still remain, however, and include metal-
Fig. 10. Coronary arteries. Three-dimensional steady-state free-precession sequence obtained during a single breath-hold
demonstrates the proximal extent of the right coronary artery, the left main coronary artery, and the circumflex coronary artery.
664
lic susceptibility artifact anteriorly from sternal wires,

and difficulty imaging the graft anastomosis site
because of cardiac motion and artifact from nearby
surgical clips.
Using a nongated three-dimensional contrast-enhanced MR angiography sequence, Brenner et al
[212] studied 85 patients with a total of 247 venous
and arterial grafts. In this study, 87% of all vein grafts
and 95% of all arterial grafts were visualized, with
a sensitivity and specificity for graft patency of
90% and 94%, respectively. Other studies have
demonstrated an overall sensitivity and specificity
for graft patency of 88% to 98% and of 72% to
100% [213 215], respectively, with the more recent
studies showing specificity closer to 100%. [216]
Although CT angiography is also excellent in
assessing vein graft patency noninvasively, it cannot
quantitate flow as has recently been accurately characterized and validated by CMR [217 221]. Langerak et al [218] used a high-resolution sequence to
evaluate 71 patients with prior coronary artery bypass
grafts who presented with chest pain. When compared with cardiac catheterization, sensitivity and
specificity for detecting a greater than or equal to
70% stenosis in single-vein coronary artery bypass
grafts were 96% and 92%, respectively.
Coronary artery stenosis
Coronary MR angiography precisely characterizes
the proximal course of the coronary arteries. Overall,
MR angiography has a sensitivity approaching 100%
and specificity of 100% [176,222] for the evaluation
of coronary artery anomalies, and is increasingly
being recognized as the new standard of reference
[223 226]. The efficacy of coronary MR angiography is still limited in reliably detecting coronary
stenosis. Although evaluation of proximal and middle
segments of the coronary arteries is becoming more
reliable with current techniques, evaluation of the
distal portion of the coronary arteries is limited. For a
clinically useful complete evaluation of the coronary
arteries, however, it is necessary to evaluate the entire
coronary vascular tree. Several multicenter trials
evaluating the performance of coronary MR angiography in the detection of stenosis have been completed, with other studies currently underway. In a
multicenter trial involving 109 patients, Kim et al
[222] showed 100% sensitivity for the diagnosis of
three-vessel disease or significant left main stenosis
when the arteries were visualized. The approximate
mean length of coronary artery visualized was 2 mm
for the left main and 6, 3, and 8 mm for the left
anterior descending, circumflex, and right coronary
arteries, respectively. The sensitivity to detect a

greater than 50% lesion in each coronary artery was
88%, 83%, 72%, and 93% for the left main, left
anterior descending, circumflex, and right coronary
arteries, respectively. Smaller, more recent, trials have
demonstrated sensitivities of 75% to 98%, specificities of 75% to 100%, and diagnostic accuracies of
84% to 95% for detecting significant coronary disease [227 229]. There is tremendous clinical, research, and vendor interest in coronary MR
angiography with consistent improvement in pulse
sequence design, experience, and overcoming technical hurdles. More work is required, however, before
coronary MR angiography can replace cardiac catheterization for the complete evaluation of coronary
lumen stenosis.
Summary
Over the past two decades there has been significant progress in the field of MR imaging and its
application to the investigation of ischemic heart
disease. The concept of a single, rapid, noninvasive
examination that evaluates perfusion, morphology,
global and regional ventricular function, viability,
and coronary anatomy has been realized. Many
studies have now convincingly demonstrated the
superiority of MR imaging over other modalities for
a wide spectrum of cardiovascular disease. The
convenience of a single noninvasive test without the
limitations inherent in competing modalities ensures
CMR will become a routine diagnostic tool for
evaluating ischemic heart disease. It is expected that
MR imaging will assume a greater role in the evaluation of cardiovascular disease as the technology
becomes more clinically available, and referring
physicians and patients become more aware of and
comfortable with the modality.
References
[1] Sechtem U, Sommerhoff BA, Markiewicz W, White
RD, Cheitlin MD, Higgins CB. Regional left ventricular wall thickening by magnetic resonance imaging: evaluation in normal persons and patients with
global and regional dysfunction. Am J Cardiol 1987;
59:145 51.
[2] White HD, Norris RM, Brown MA, Brandt PW,
Whitlock RM, Wild CJ. Left ventricular end-systolic
volume as the major determinant of survival after recovery from myocardial infarction. Circulation 1987;
76:44 51.
[3] Marantz PR, Tobin JN, Wassertheil-Smoller S, et al.
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
The relationship between left ventricular systolic

function and congestive heart failure diagnosed by
clinical criteria. Circulation 1988;77:607 12.
Bellenger NG, Davies LC, Francis JM, Coats AJ,
Pennell DJ. Reduction in sample size for studies of
remodeling in heart failure by the use of cardiovascular magnetic resonance. J Cardiovasc Magn Reson
2000;2:271 8.
Cerqueira MD, Weissman NJ, Dilsizian V, et al. Standardized Myocardial Segmentation and Nomenclature
for Tomographic Imaging of the Heart: a statement
for healthcare professionals from the Cardiac Imaging
Committee of the Council on Clinical Cardiology of
the American Heart Association. Circulation 2002;
105:539 42.
Rehr R, Malloy C, Filipchuk N, Peshock R. Left
ventricular volumes measured by MR imaging. Radiology 1985;156:717 9.
Sakuma H, Globits S, Bourne MW, Shimakawa A,
Foo TK, Higgins CB. Improved reproducibility in
measuring LV volumes and mass using multicoil
breath-hold cine MR imaging. J Magn Reson Imaging
1996;6:124 7.
Forbat SM, Sakrana MA, Darasz KH, El-Demerdash
F, Underwood SR. Rapid assessment of left ventricular volume by short axis cine MRI. Br J Radiol 1996;
69:221 5.
Bellenger NG, Grothues F, Smith GC, Pennell DJ.
Quantification of right and left ventricular function
by cardiovascular magnetic resonance. Herz 2000;
25:392 9.
Gopal AS, Shen Z, Sapin PM, et al. Assessment of
cardiac function by three-dimensional echocardiography compared with conventional noninvasive methods. Circulation 1995;92:842 53.
Danias PG, Chuang ML, Parker RA, et al. Relation
between the number of image planes and the accuracy
of three-dimensional echocardiography for measuring
left ventricular volumes and ejection fraction. Am J
Cardiol 1998;82:1431 4.
Thiele H, Nagel E, Paetsch I, et al. Functional cardiac
MR imaging with steady-state free precession (SSFP)
significantly improves endocardial border delineation
without contrast agents. J Magn Reson Imaging 2001;
14:362 7.
Li W, Stern JS, Mai VM, Pierchala LN, Edelman RR,
Prasad PV. MR assessment of left ventricular function:
quantitative comparison of fast imaging employing
steady-state acquisition (FIESTA) with fast gradient
echo cine technique. J Magn Reson Imaging 2002;
16:559 64.
Thiele H, Paetsch I, Schnackenburg B, et al. Improved accuracy of quantitative assessment of left
ventricular volume and ejection fraction by geometric
models with steady-state free precession. J Cardiovasc Magn Reson 2002;4:327 39.
Barkhausen J, Ruehm SG, Goyen M, Buck T, Laub
G, Debatin JF. MR evaluation of ventricular function:
true fast imaging with steady-state precession versus
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
665
fast low-angle shot cine MR imaging: feasibility

study. Radiology 2001;219:264 9.
Plein S, Bloomer TN, Ridgway JP, Jones TR, Bainbridge GJ, Sivananthan MU. Steady-state free precession magnetic resonance imaging of the heart:
comparison with segmented k-space gradient-echo
imaging. J Magn Reson Imaging 2001;14:230 6.
Castillo E, Lima JA, Bluemke DA. Regional myocardial function: advances in MR imaging and analysis.
Radiographics 2003;23(Spec No):S127 40.
Hori Y, Yamada N, Higashi M, Hirai N, Nakatani S.
Rapid evaluation of right and left ventricular function
and mass using real-time true-FISP cine MR imaging
without breath-hold: comparison with segmented trueFISP cine MR imaging with breath-hold. J Cardiovasc
Magn Reson 2003;5:439 50.
Zerhouni EA, Parish DM, Rogers WJ, Yang A, Shapiro EP. Human heart: tagging with MR imaging a
method for noninvasive assessment of myocardial
motion. Radiology 1988;169:59 63.
Buchalter MB, Weiss JL, Rogers WJ, et al. Noninvasive quantification of left ventricular rotational deformation in normal humans using magnetic resonance
imaging myocardial tagging. Circulation 1990;81:
1236 44.
Bolster Jr BD, McVeigh ER, Zerhouni EA. Myocardial tagging in polar coordinates with use of striped
tags. Radiology 1990;177:769 72.
Lima JA, Jeremy R, Guier W, et al. Accurate systolic
wall thickening by nuclear magnetic resonance imaging with tissue tagging: correlation with sonomicrometers in normal and ischemic myocardium. J Am Coll
Cardiol 1993;21:1741 51.
Croisille P, Moore CC, Judd RM, et al. Differentiation of viable and nonviable myocardium by the
use of three-dimensional tagged MRI in 2-day-old
reperfused canine infarcts. Circulation 1999;99:
284 91.
Ryf S, Spiegel MA, Gerber M, Boesiger P. Myocardial tagging with 3D-CSPAMM. J Magn Reson Imaging 2002;16:320 5.
Kramer CM, Malkowski MJ, Mankad S, Theobald
TM, Pakstis DL, Rogers Jr WJ. Magnetic resonance tagging and echocardiographic response to
dobutamine and functional improvement after reperfused myocardial infarction. Am Heart J 2002;143:
1046 51.
Dubach P, Myers J, Bonetti P, et al. Effects of bisoprolol fumarate on left ventricular size, function, and
exercise capacity in patients with heart failure: analysis with magnetic resonance myocardial tagging.
Am Heart J 2002;143:676 83.
Sandstede JJ, Johnson T, Harre K, et al. Cardiac systolic rotation and contraction before and after valve
replacement for aortic stenosis: a myocardial tagging
study using MR imaging. AJR Am J Roentgenol
2002;178:953 8.
Ennis DB, Epstein FH, Kellman P, Fananapazir L,
McVeigh ER, Arai AE. Assessment of regional sys-
666
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]

tolic and diastolic dysfunction in familial hypertrophic
cardiomyopathy using MR tagging. Magn Reson Med
2003;50:638 42.
Wedeen VJ. Magnetic resonance imaging of myocardial kinematics: technique to detect, localize, and
quantify the strain rates of the active human myocardium. Magn Reson Med 1992;27:52 67.
Osman NF, Kerwin WS, McVeigh ER, Prince JL.
Cardiac motion tracking using CINE harmonic phase
(HARP) magnetic resonance imaging. Magn Reson
Med 1999;42:1048 60.
Osman NF, Sampath S, Atalar E, Prince JL. Imaging
longitudinal cardiac strain on short-axis images using
strain-encoded MRI. Magn Reson Med 2001;46:
324 34.
Arai AE, Gaither III CC, Epstein FH, Balaban RS,
Wolff SD. Myocardial velocity gradient imaging by
phase contrast MRI with application to regional function in myocardial ischemia. Magn Reson Med 1999;
42:98 109.
Reese TG, Feinberg DA, Dou J, Wedeen VJ. Phase
contrast MRI of myocardial 3D strain by encoding
contiguous slices in a single shot. Magn Reson Med
2002;47:665 76.
Aletras AH, Balaban RS, Wen H. High-resolution
strain analysis of the human heart with fast-DENSE.
J Magn Reson 1999;140:41 57.
Aletras AH, Ding S, Balaban RS, Wen H. DENSE:
displacement encoding with stimulated echoes in
cardiac functional MRI. J Magn Reson 1999;137:
247 52.
Aletras AH, Wen H. Mixed echo train acquisition
displacement encoding with stimulated echoes: an
optimized DENSE method for in vivo functional
imaging of the human heart. Magn Reson Med
2001;46:523 34.
McVeigh ER. MRI of myocardial function: motion
tracking techniques. Magn Reson Imaging 1996;14:
137 50.
Klocke FJ, Simonetti OP, Judd RM, et al. Limits
of detection of regional differences in vasodilated
flow in viable myocardium by first-pass magnetic
resonance perfusion imaging. Circulation 2001;104:
2412 6.
Wilke N, Simm C, Zhang J, et al. Contrast-enhanced
first pass myocardial perfusion imaging: correlation
between myocardial blood flow in dogs at rest
and during hyperemia. Magn Reson Med 1993;29:
485 97.
Atkinson DJ, Burstein D, Edelman RR. First-pass
cardiac perfusion: evaluation with ultrafast MR imaging. Radiology 1990;174(3 Pt 1):757 62.
Wilke N, Machnig T, Engels G. Dynamic perfusion
studies by ultrafast MR imaging: initial clinical
results from cardiology. Electromedica 1990;58:
102 8.
Manning WJ, Atkinson DJ, Grossman W, Paulin S,
Edelman RR. First-pass nuclear magnetic resonance
imaging studies using gadolinium-DTPA in patients
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
with coronary artery disease. J Am Coll Cardiol

1991;18:959 65.
Klein MA, Collier BD, Hellman RS, Bamrah VS.
Detection of chronic coronary artery disease: value
of pharmacologically stressed, dynamically enhanced
turbo-fast low-angle shot MR images. AJR Am J
Roentgenol 1993;161:257 63.
Eichenberger AC, Schuiki E, Kochli VD, Amann FW,
McKinnon GC, von Schulthess GK. Ischemic heart
disease: assessment with gadolinium-enhanced ultrafast MR imaging and dipyridamole stress. J Magn
Hartnell G, Cerel A, Kamalesh M, et al. Detection of
myocardial ischemia: value of combined myocardial
perfusion and cineangiographic MR imaging. AJR
Saeed M, Wendland MF, Lauerma K, et al. First-pass
contrast-enhanced inversion recovery and driven equilibrium fast GRE imaging studies: detection of acute
myocardial ischemia. J Magn Reson Imaging 1995;
5:515 23.
Kraitchman DL, Wilke N, Hexeberg E, et al. Myocardial perfusion and function in dogs with moderate coronary stenosis. Magn Reson Med 1996;35:
771 80.
Wilke N, Jerosch-Herold M, Wang Y, et al. Myocardial perfusion reserve: assessment with multisection,
quantitative, first-pass MR imaging. Radiology 1997;
204:373 84.
Braunwald E. Heart disease: a textbook of cardiovascular medicine, vol. 1. 5th edition. Philadelphia:
WB Saunders; 1997.
Marcus ML, Wilson RF, White CW. Methods of
measurement of myocardial blood flow in patients:
a critical review. Circulation 1987;76:245 53.
Gould KL, Lipscomb K. Effects of coronary stenoses
on coronary flow reserve and resistance. Am J Cardiol 1974;34:48 55.
Wilson RF, Marcus ML, White CW. Prediction of the
physiologic significance of coronary arterial lesions
by quantitative lesion geometry in patients with limited coronary artery disease. Circulation 1987;75:
723 32.
Wilson RF, Laughlin DE, Ackell PH, et al. Transluminal, subselective measurement of coronary artery
blood flow velocity and vasodilator reserve in man.
Schreiber WG, Schmitt M, Kalden P, Mohrs OK,
Kreitner KF, Thelen M. Dynamic contrast-enhanced
myocardial perfusion imaging using saturation-prepared TrueFISP. J Magn Reson Imaging 2002;16:
641 52.
Lima JA, Judd RM, Bazille A, Schulman SP, Atalar E,
Zerhouni EA. Regional heterogeneity of human myocardial infarcts demonstrated by contrast-enhanced
MRI: potential mechanisms. Circulation 1995;92:
1117 25.
Rochitte CE, Lima JA, Bluemke DA, et al. Magnitude and time course of microvascular obstruction
1167
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
and tissue injury after acute myocardial infarction.

Circulation 1998;98:1006 14.
Kim RJ, Chen EL, Lima JA, Judd RM. Myocardial
Gd-DTPA kinetics determine MRI contrast enhancement and reflect the extent and severity of myocardial
injury after acute reperfused infarction. Circulation
1996;94:3318 26.
Saeed M, Wendland MF, Yu KK, et al. Identification
of myocardial reperfusion with echo planar magnetic
resonance imaging: discrimination between occlusive
and reperfused infarctions. Circulation 1994;90:
1492 501.
Saeed M, Wendland MF, Takehara Y, Masui T, Higgins CB. Reperfusion and irreversible myocardial
injury: identification with a nonionic MR imaging
contrast medium. Radiology 1992;182:675 83.
Rogers Jr WJ, Kramer CM, Geskin G, et al. Early
contrast-enhanced MRI predicts late functional recovery after reperfused myocardial infarction. Circulation
1999;99:744 50.
Wu KC, Zerhouni EA, Judd RM, et al. Prognostic
significance of microvascular obstruction by magnetic resonance imaging in patients with acute myocardial infarction. Circulation 1998;97:765 72.
Chiu CW, So NM, Lam WW, Chan KY, Sanderson
JE. Combined first-pass perfusion and viability study
at MR imaging in patients with non-ST segmentelevation acute coronary syndromes: feasibility study.
Radiology 2003;226:717 22.
Sandstede JJ, Lipke C, Beer M, et al. Analysis of
first-pass and delayed contrast-enhancement patterns
of dysfunctional myocardium on MR imaging: use in
the prediction of myocardial viability. AJR Am J
Roentgenol 2000;174:1737 40.
Lauerma K, Niemi P, Hanninen H, et al. Multimodality MR imaging assessment of myocardial viability:
combination of first-pass and late contrast enhancement to wall motion dynamics and comparison with
FDG PET-initial experience. Radiology 2000;217:
729 36.
Kwong RY, Schussheim AE, Rekhraj S, et al. Detecting acute coronary syndrome in the emergency department with cardiac magnetic resonance imaging.
Ishida N, Sakuma H, Motoyasu M, et al. Noninfarcted
myocardium: correlation between dynamic first-pass
contrast-enhanced myocardial MR imaging and quantitative coronary angiography. Radiology 2003;229:
209 16.
Laddis T, Manning WJ, Danias PG. Cardiac MRI for
assessment of myocardial perfusion: current status and
future perspectives. J Nucl Cardiol 2001;8:207 14.
Matheijssen NA, Louwerenburg HW, van Rugge FP,
et al. Comparison of ultrafast dipyridamole magnetic
resonance imaging with dipyridamole SestaMIBI
SPECT for detection of perfusion abnormalities in
patients with one-vessel coronary artery disease: assessment by quantitative model fitting. Magn Reson
Med 1996;35:221 8.
667
[69] Al-Saadi N, Nagel E, Gross M, et al. Noninvasive

detection of myocardial ischemia from perfusion reserve based on cardiovascular magnetic resonance.
Circulation 2000;101:1379 83.
[70] Nagel E, Klein C, Paetsch I, et al. Magnetic resonance
perfusion measurements for the noninvasive detection of coronary artery disease. Circulation 2003;
108:432 7.
[71] Cullen JH, Horsfield MA, Reek CR, Cherryman GR,
Barnett DB, Samani NJ. A myocardial perfusion
reserve index in humans using first-pass contrastenhanced magnetic resonance imaging. J Am Coll
Cardiol 1999;33:1386 94.
[72] Sensky PR, Jivan A, Hudson NM, et al. Coronary
artery disease: combined stress MR imaging protocol-one-stop evaluation of myocardial perfusion and
function. Radiology 2000;215:608 14.
[73] Keijer JT, van Rossum AC, Wilke N, et al. Magnetic
resonance imaging of myocardial perfusion in singlevessel coronary artery disease: implications for
transmural assessment of myocardial perfusion. J
Cardiovasc Magn Reson 2000;2:189 200.
[74] Keijer JT, van Rossum AC, van Eenige MJ, et al.
Magnetic resonance imaging of regional myocardial
perfusion in patients with single-vessel coronary artery disease: quantitative comparison with (201)Thallium-SPECT and coronary angiography. J Magn
[75] Schwitter J, Nanz D, Kneifel S, et al. Assessment of
myocardial perfusion in coronary artery disease by
magnetic resonance: a comparison with positron
emission tomography and coronary angiography. Circulation 2001;103:2230 5.
[76] Ibrahim T, Nekolla SG, Schreiber K, et al. Assessment
of coronary flow reserve: comparison between contrast-enhanced magnetic resonance imaging and positron emission tomography. J Am Coll Cardiol 2002;
39:864 70.
[77] Pennell DJ, Underwood SR, Manzara CC, et al. Magnetic resonance imaging during dobutamine stress in
coronary artery disease. Am J Cardiol 1992;70:34 40.
[78] Rerkpattanapipat P, Gandhi SK, Darty SN, et al. Feasibility to detect severe coronary artery stenoses with
upright treadmill exercise magnetic resonance imaging. Am J Cardiol 2003;92:603 6.
[79] Hundley WG, Hamilton CA, Thomas MS, et al. Utility of fast cine magnetic resonance imaging and
display for the detection of myocardial ischemia in
patients not well suited for second harmonic stress
echocardiography. Circulation 1999;100:1697 702.
[80] Hundley WG, Morgan TM, Neagle CM, Hamilton
CA, Rerkpattanapipat P, Link KM. Magnetic resonance imaging determination of cardiac prognosis.
Circulation 2002;106:2328 33.
[81] Nagel E, Lehmkuhl HB, Bocksch W, et al. Noninvasive diagnosis of ischemia-induced wall motion
abnormalities with the use of high-dose dobutamine
stress MR1: comparison with dobutamine stress echocardiography. Circulation 1999;99:763 70.
668
[82] Bauer WR, Nadler W, Bock M, et al. Theory of the

BOLD effect in the capillary region: an analytical
approach for the determination of T2 in the capillary
network of myocardium. Magn Reson Med 1999;41:
51 62.
[83] Reeder SB, Holmes AA, McVeigh ER, Forder JR.
Simultaneous noninvasive determination of regional
myocardial perfusion and oxygen content in rabbits:
toward direct measurement of myocardial oxygen
consumption at MR imaging. Radiology 1999;212:
739 47.
[84] Niemi P, Poncelet BP, Kwong KK, et al. Myocardial
intensity changes associated with flow stimulation
in blood oxygenation sensitive magnetic resonance
imaging. Magn Reson Med 1996;36:78 82.
[85] Reeder SB, Faranesh AZ, Boxerman JL, McVeigh
ER. In vivo measurement of T*2 and field inhomogeneity maps in the human heart at 1.5 T. Magn
Reson Med 1998;39:988 98.
[86] Wacker CM, Bock M, Hartlep AW, et al. BOLD-MRI
in ten patients with coronary artery disease: evidence
for imaging of capillary recruitment in myocardium
supplied by the stenotic artery. Magma 1999;8:48 54.
[87] Beache GM, Herzka DA, Boxerman JL, et al. Attenuated myocardial vasodilator response in patients with
hypertensive hypertrophy revealed by oxygenationdependent magnetic resonance imaging. Circulation
2001;104:1214 7.
[88] Anderson LJ, Holden S, Davis B, et al. Cardiovascular
T2-star (T2*) magnetic resonance for the early diagnosis of myocardial iron overload. Eur Heart J 2001;
22:2171 9.
[89] Wacker CM, Hartlep AW, Pfleger S, Schad LR, Ertl
G, Bauer WR. Susceptibility-sensitive magnetic
resonance imaging detects human myocardium supplied by a stenotic coronary artery without a contrast
agent. J Am Coll Cardiol 2003;41:834 40.
[90] Rahimtoola SH. Hibernating myocardium has reduced
blood flow at rest that increases with low-dose dobutamine. Circulation 1996;94:3055 61.
[91] Rahimtoola SH. The hibernating myocardium. Am
Heart J 1989;117:211 21.
[92] Braunwald E, Rutherford JD. Reversible ischemic
left ventricular dysfunction: evidence for the hibernating myocardium. J Am Coll Cardiol 1986;8:
1467 70.
[93] Metcalfe MJ, Cross S, Norton MY, Jennings K, Walton S. Objective measurement of changes in regional
wall motion characteristics following coronary artery
bypass grafting using tomographic radionuclide ventriculography. Nucl Med Commun 1995;16:539 47.
[94] Rahimtoola SH. A perspective on the three large multicenter randomized clinical trials of coronary bypass
surgery for chronic stable angina. Circulation 1985;
72(6 Pt 2):V123 35.
[95] Tillisch J, Brunken R, Marshall R, et al. Reversibility
of cardiac wall-motion abnormalities predicted by
positron tomography. N Engl J Med 1986;314:884 8.
[96] Dilsizian V, Rocco TP, Freedman NM, Leon MB,
[97]
[98]
[99]
[100]
[101]
[102]
[103]
[104]
[105]
[106]
[107]
Bonow RO. Enhanced detection of ischemic but viable myocardium by the reinjection of thallium after
stress-redistribution imaging. N Engl J Med 1990;
323:141 6.
Ragosta M, Beller GA, Watson DD, Kaul S, Gimple
LW. Quantitative planar rest-redistribution 201Tl
imaging in detection of myocardial viability and prediction of improvement in left ventricular function
after coronary bypass surgery in patients with severely depressed left ventricular function. Circulation
1993;87:1630 41.
Arnese M, Cornel JH, Salustri A, et al. Prediction of
improvement of regional left ventricular function after
surgical revascularization: a comparison of low-dose
dobutamine echocardiography with 201Tl single-photon emission computed tomography. Circulation 1995;
91:2748 52.
Pagley PR, Beller GA, Watson DD, Gimple LW,
Ragosta M. Improved outcome after coronary bypass
surgery in patients with ischemic cardiomyopathy
and residual myocardial viability. Circulation 1997;
96:793 800.
Haas F, Haehnel CJ, Picker W, et al. Preoperative
positron emission tomographic viability assessment
and perioperative and postoperative risk in patients
with advanced ischemic heart disease. J Am Coll
Cardiol 1997;30:1693 700.
Di Carli MF, Maddahi J, Rokhsar S, et al. Long-term
survival of patients with coronary artery disease and
left ventricular dysfunction: implications for the
role of myocardial viability assessment in management decisions. J Thorac Cardiovasc Surg 1998;116:
997 1004.
Ho KK, Anderson KM, Kannel WB, Grossman W,
Levy D. Survival after the onset of congestive heart
failure in Framingham Heart Study subjects. Circulation 1993;88:107 15.
Alderman EL, Fisher LD, Litwin P, et al. Results of
coronary artery surgery in patients with poor left
ventricular function (CASS). Circulation 1983;68:
785 95.
Arheden H, Saeed M, Higgins CB, et al. Measurement of the distribution volume of gadopentetate
dimeglumine at echo-planar MR imaging to quantify
myocardial infarction: comparison with 99mTcDTPA autoradiography in rats. Radiology 1999;211:
698 708.
Arheden H, Saeed M, Higgins CB, et al. Reperfused
rat myocardium subjected to various durations of
ischemia: estimation of the distribution volume of
contrast material with echo-planar MR imaging.
Radiology 2000;215:520 8.
Lima JA. Myocardial viability: is MRI contrast enhancement the new gold standard? Presented at
Transcatheter Therapeutics. Washington, DC, September 18 21, 2003.
Simonetti OP, Kim RJ, Fieno DS, et al. An improved
MR imaging technique for the visualization of myocardial infarction. Radiology 2001;218:215 23.

[108] Inoue S, Shimada T, Murakami Y. Clinical significance of gadolinium-DTPA-enhanced MRI for
detection of myocardial lesions in a patient with sarcoidosis. Clin Radiol 1999;54:70 2.
[109] Matsuki M, Matsuo M. MR findings of myocardial
sarcoidosis. Clin Radiol 2000;55:323 5.
[110] Shimada T, Shimada K, Sakane T, et al. Diagnosis of
cardiac sarcoidosis and evaluation of the effects of
steroid therapy by gadolinium-DTPA-enhanced magnetic resonance imaging. Am J Med 2001;110:520 7.
[111] Vignaux O, Dhote R, Duboc D, et al. Detection of
myocardial involvement in patients with sarcoidosis
applying T2-weighted, contrast-enhanced, and cine
magnetic resonance imaging: initial results of a prospective study. J Comput Assist Tomogr 2002;26:
762 7.
[112] Vignaux O, Dhote R, Duboc D, et al. Clinical significance of myocardial magnetic resonance abnormalities in patients with sarcoidosis: a 1-year follow-up
study. Chest 2002;122:1895 901.
[113] Dhote R, Vignaux O, Blanche P, et al. Value of MRI
for the diagnosis of cardiac involvement in sarcoidosis]. Rev Med Interne 2003;24:151 7.
[114] Serra JJ, Monte GU, Mello ES, et al. Images in
cardiovascular medicine: cardiac sarcoidosis evaluated by delayed-enhanced magnetic resonance imaging. Circulation 2003;107:e188 9.
[115] Friedrich MG, Strohm O, Schulz-Menger J, Marciniak H, Luft FC, Dietz R. Contrast media-enhanced
magnetic resonance imaging visualizes myocardial
changes in the course of viral myocarditis. Circulation
1998;97:1802 9.
[116] Laissy J-P, Messin B, Varenne O, et al. MRI of acute
myocarditis: a comprehensive approach based on
various imaging sequences. Chest 2002;122:1638 48.
[117] Choudhury L, Mahrholdt H, Wagner A, et al. Myocardial scarring in asymptomatic or mildly symptomatic patients with hypertrophic cardiomyopathy.
[118] Kim RJ, Judd RM. Gadolinium-enhanced magnetic
resonance imaging in hypertrophic cardiomyopathy:
in vivo imaging of the pathologic substrate for premature cardiac death? J Am Coll Cardiol 2003;41:
1568 72.
[119] Moon J, McKenna W, McCrohon J, Elliott P, Smith G,
Pennell D. Towards clinical risk assessment in hypertrophic cardiomyopathy with gadolinium cardiovascular magnetic resonance. J Am Coll Cardiol 2003;
41:1561 7.
[120] Wilson JM, Villareal RP, Hariharan R, Massumi A,
Muthupillai R, Flamm SD. Magnetic resonance imaging of myocardial fibrosis in hypertrophic cardiomyopathy. Tex Heart Inst J 2002;29:176 80.
[121] Wesbey GE, Higgins CB, McNamara MT, et al. Effect of gadolinium-DTPA on the magnetic relaxation
times of normal and infarcted myocardium. Radiology 1984;153:165 9.
[122] McNamara MT, Tscholakoff D, Revel D, et al. Differentiation of reversible and irreversible myocardial
[123]
[124]
[125]
[126]
[127]
[128]
[129]
[130]
[131]
[132]
[133]
[134]
[135]
[136]
669
injury by MR imaging with and without gadoliniumDTPA. Radiology 1986;158:765 9.

Schaefer S, Malloy CR, Katz J, et al. GadoliniumDTPA-enhanced nuclear magnetic resonance imaging of reperfused myocardium: identification of the
myocardial bed at risk. J Am Coll Cardiol 1988;12:
1064 72.
Saeed M, Wendland MF, Masui T, Higgins CB.
Reperfused myocardial infarctions on T1- and susceptibility-enhanced MRI: evidence for loss of compartmentalization of contrast media. Magn Reson
Med 1994;31:31 9.
Judd RM, Lugo-Olivieri CH, Arai M, et al. Physiological basis of myocardial contrast enhancement in
fast magnetic resonance images of 2-day-old reperfused canine infarcts. Circulation 1995;92:1902 10.
Choi KM, Kim RJ, Gubernikoff G, Vargas JD, Parker
M, Judd RM. Transmural extent of acute myocardial
infarction predicts long-term improvement in contractile function. Circulation 2001;104:1101 7.
Gerber BL, Garot J, Bluemke DA, Wu KC, Lima
JAC. Accuracy of contrast-enhanced magnetic resonance imaging in predicting improvement of regional
myocardial function in patients after acute myocardial
infarction. Circulation 2002;106:1083 9.
Kim RJ, Fieno DS, Parrish TB, et al. Relationship
of MRI delayed contrast enhancement to irreversible
injury, infarct age, and contractile function. Circulation 1999;100:1992 2002.
Beek AM, Kuhl HP, Bondarenko O, et al. Delayed
contrast-enhanced magnetic resonance imaging for
the prediction of regional functional improvement
after acute myocardial infarction. J Am Coll Cardiol
2003;42:895 901.
Kim RJ, Wu E, Rafael A, et al. The use of contrastenhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 2000;
343:1445 53.
Wu E, Judd RM, Vargas JD, Klocke FJ, Bonow RO,
Kim RJ. Visualisation of presence, location, and
transmural extent of healed Q-wave and non Q-wave
myocardial infarction. Lancet 2001;357:21 8.
Bache RJ, Schwartz JS. Effect of perfusion pressure
distal to a coronary stenosis on transmural myocardial
blood flow. Circulation 1982;65:928 35.
Goldberg RJ, Gore JM, Alpert JS, Dalen JE. Non Q
wave myocardial infarction: recent changes in occurrence and prognosisa community-wide perspective.
Am Heart J 1987;113(2 Pt 1):273 9.
Nicod P, Gilpin E, Dittrich H, et al. Short- and longterm clinical outcome after Q wave and non Q wave
myocardial infarction in a large patient population.
Circulation Mar 1989;79(3):528 36.
Aguirre FV, Younis LT, Chaitman BR, et al. Early and
1-year clinical outcome of patients evolving non Qwave versus Q-wave myocardial infarction after
thrombolysis: results from the TIMI II Study. Circulation 1995;91:2541 8.
Herlitz J, Karlson BW, Sjolin M, Lindqvist J. Ten
670
[137]
[138]
[139]
[140]
[141]
[142]
[143]
[144]
[145]
[146]
[147]
[148]

year mortality in subsets of patients with an acute
coronary syndrome. Heart 2001;86:391 6.
Pauletto P, Piccolo D, Scannapieco G, et al. Changes
in myoglobin, creatine kinase and creatine kinase-MB
after percutaneous transluminal coronary angioplasty
for stable angina pectoris. Am J Cardiol 1987;59:
999 1000.
Harrington RA, Lincoff AM, Califf RM, et al. Characteristics and consequences of myocardial infarction
after percutaneous coronary intervention: insights
from the Coronary Angioplasty Versus Excisional
Atherectomy Trial (CAVEAT). J Am Coll Cardiol
1995;25:1693 9.
Abdelmeguid AE, Ellis SG, Sapp SK, Whitlow PL,
Topol EJ. Defining the appropriate threshold of
creatine kinase elevation after percutaneous coronary
interventions. Am Heart J 1996;131:1097 105.
Califf RM, Abdelmeguid AE, Kuntz RE, et al. Myonecrosis after revascularization procedures. J Am Coll
Cardiol 1998;31:241 51.
Tardiff BE, Califf RM, Tcheng JE, et al. Clinical outcomes after detection of elevated cardiac enzymes
in patients undergoing percutaneous intervention.
IMPACT-II Investigators. Integrilin (eptifibatide) to
Minimize Platelet Aggregation and Coronary Thrombosis-II. J Am Coll Cardiol 1999;33:88 96.
Abdelmeguid AE, Topol EJ, Whitlow PL, Sapp SK,
Ellis SG. Significance of mild transient release of
creatine kinase-MB fraction after percutaneous coronary interventions. Circulation 1996;94:1528 36.
Chida K, Sugiura M, Ohkawa S, et al. A clinicopathologic correlation study of thallium-201 myocardial
scintigraphy in diagnosis of myocardial infarction.
Jpn Heart J 1987;28:307 21.
Kontos MC, Kurdziel KA, Ornato JP, Schmidt KL,
Jesse RL, Tatum JL. A nonischemic electrocardiogram does not always predict a small myocardial
infarction: results with acute myocardial perfusion
imaging. Am Heart J 2001;141:360 6.
Miller TD, Christian TF, Hopfenspirger MR, Hodge
DO, Gersh BJ, Gibbons RJ. Infarct size after acute
myocardial infarction measured by quantitative tomographic 99mTc sestamibi imaging predicts subsequent mortality. Circulation 1995;92:334 41.
Wagner A, Mahrholdt H, Holly TA, et al. Contrastenhanced MRI and routine single photon emission
computed tomography (SPECT) perfusion imaging
for detection of subendocardial myocardial infarcts:
an imaging study. Lancet 2003;361:374 9.
Kannel WB, Abbott RD. Incidence and prognosis of
unrecognized myocardial infarction: an update on the
Framingham study. N Engl J Med 1984;311:1144 7.
Kitagawa K, Sakuma H, Hirano T, Okamoto S,
Makino K, Takeda K. Acute myocardial infarction: myocardial viability assessment in patients
early thereafter comparison of contrast-enhanced
MR imaging with resting (201)Tl SPECT. Single
photon emission computed tomography. Radiology
2003;226:138 44.
[149] Bax JJ, Wijns W, Cornel JH, Visser FC, Boersma E,

Fioretti PM. Accuracy of currently available techniques for prediction of functional recovery after
revascularization in patients with left ventricular dysfunction due to chronic coronary artery disease: comparison of pooled data. J Am Coll Cardiol 1997;30:
1451 60.
[150] Bax JJ, Poldermans D, Elhendy A, Boersma E,
Rahimtoola SH. Sensitivity, specificity, and predictive accuracies of various noninvasive techniques for
detecting hibernating myocardium. Curr Probl Cardiol 2001;26:141 86.
[151] Klein C, Nekolla SG, Bengel FM, et al. Assessment
of myocardial viability with contrast-enhanced magnetic resonance imaging: comparison with positron
emission tomography. Circulation 2002;105:162 7.
[152] Bonow RO. Myocardial viability and prognosis in
patients with ischemic left ventricular dysfunction.
[153] Bax JJ, Fath-Ordoubadi F, Boersma E, Wijns W,
Camici PG. Accuracy of PET in predicting functional
recovery after revascularisation in patients with
chronic ischaemic dysfunction: head-to-head comparison between blood flow, glucose utilisation and
water-perfusable tissue fraction. Eur J Nucl Med Mol
Imaging 2002;29:721 7.
[154] Kuhl HP, Beek AM, van der Weerdt AP, et al. Myocardial viability in chronic ischemic heart disease:
comparison of contrast-enhanced magnetic resonance
imaging with (18)F-fluorodeoxyglucose positron
emission tomography. J Am Coll Cardiol 2003;41:
1341 8.
[155] Gunning MG, Kaprielian RR, Pepper J, et al. The
histology of viable and hibernating myocardium in
relation to imaging characteristics. J Am Coll Cardiol
2002;39:428 35.
[156] Sandstede JJ, Bertsch G, Beer M, et al. Detection of
myocardial viability by low-dose dobutamine cine MR
imaging. Magn Reson Imaging 1999;17: 1437 43.
[157] Baer FM, Voth E, Schneider CA, Theissen P, Schicha
H, Sechtem U. Comparison of low-dose dobutamine-gradient-echo magnetic resonance imaging and
positron emission tomography with [18F]fluorodeoxyglucose in patients with chronic coronary artery
disease: a functional and morphological approach to
the detection of residual myocardial viability. Circulation 1995;91:1006 15.
[158] Mollet NR, Dymarkowski S, Volders W, et al. Visualization of ventricular thrombi with contrast-enhanced
magnetic resonance imaging in patients with ischemic
heart disease. Circulation 2002;106:2873 6.
[159] Barkhausen J, Hunold P, Eggebrecht H, et al. Detection and characterization of intracardiac thrombi
on MR imaging. AJR Am J Roentgenol 2002;179:
1539 44.
[160] Paydarfar D, Krieger D, Dib N, et al. In vivo magnetic resonance imaging and surgical histopathology
of intracardiac masses: distinct features of subacute
thrombi. Cardiology 2001;95:40 7.

[161] Wang Y, Riederer SJ, Ehman RL. Respiratory motion
of the heart: kinematics and the implications for the
spatial resolution in coronary imaging. Magn Reson
Med 1995;33:713 9.
[162] Steenbeck J, Pruessmann K. Technical developments
in cardiac MRI: 2000 update. Rays 2001;26:15 34.
[163] Bell M. Software eliminates need for ECG during
cardiac MR exams. Available at: http://www.siemens.
co.uk/med/news/current/CardiaMRI0203.pdf.
Accessed February 21, 2003.
[164] Stuber M, Botnar RM, Danias PG, Kissinger KV,
Manning WJ. Breathhold three-dimensional coronary
magnetic resonance angiography using real-time
navigator technology. J Cardiovasc Magn Reson
1999;1:233 8.
[165] Stuber M, Danias PG, Botnar RM, Sodickson DK,
Kissinger KV, Manning WJ. Superiority of prone
position in free-breathing 3D coronary MRA in
patients with coronary disease. J Magn Reson Imaging 2001;13:185 91.
[166] Liu YL, Riederer SJ, Rossman PJ, Grimm RC, Debbins JP, Ehman RL. A monitoring, feedback, and
triggering system for reproducible breath-hold MR
imaging. Magn Reson Med 1993;30:507 11.
[167] Doyle M, Scheidegger MB, de Graaf RG, Vermeulen J, Pohost GM. Coronary artery imaging in multiple 1-sec breath holds. Magn Reson Imaging 1993;
11:3 6.
[168] Huber ME, Oelhafen ME, Kozerke S, Weber OM,
Boesiger P. Single breath-hold extended free-breathing navigator-gated three-dimensional coronary MRA.
J Magn Reson Imaging 2002;15:210 4.
[169] Weiger M, Bornert P, Proksa R, Schaffter T, Haase A.
Motion-adapted gating based on k-space weighting
for reduction of respiratory motion artifacts. Magn
Reson Med 1997;38:322 33.
[170] Sinkus R, Bornert P. Motion pattern adapted realtime respiratory gating. Magn Reson Med 1999;41:
148 55.
[171] Weber C, Steiner P, Sinkus R, Dill T, Bornert P, Adam
G. Correlation of 3D MR coronary angiography with
selective coronary angiography: feasibility of the
motion-adapted gating technique. Eur Radiol 2002;
12:718 26.
[172] Sachs TS, Meyer CH, Irarrazabal P, Hu BS, Nishimura DG, Macovski A. The diminishing variance
algorithm for real-time reduction of motion artifacts
in MRI. Magn Reson Med 1995;34:412 22.
[173] Sachs TS, Meyer CH, Pauly JM, Hu BS, Nishimura
DG, Macovski A. The real-time interactive 3-D-DVA
for robust coronary MRA. IEEE Trans Med Imaging
2000;19:73 9.
[174] Duerinckx AJ, Atkinson DP, Mintorovitch J, Simonetti OP, Vrman MK. Two-dimensional coronary
MRA: limitations and artifacts. Eur Radiol 1996;6:
312 25.
[175] Slavin GS, Riederer SJ, Ehman RL. Two-dimensional
multishot echo-planar coronary MR angiography.
Magn Reson Med 1998;40:883 9.
671
[176] Dirksen MS, Lamb HJ, Doornhos J, Bax J. Coronary

magnetic resonance angiography: technical developments and clinical applications. J Cardiovasc Magn
Reson 2003;5:365 86.
[177] Etienne A, Botnar RM, Van Muiswinkel AM, Boesiger P, Manning WJ, Stuber M. Soap-bubble visualization and quantitative analysis of 3D coronary
magnetic resonance angiograms. Magn Reson Med
2002;48:658 66.
[178] Bornert P, Stuber M, Botnar RM, et al. Direct comparison of 3D spiral vs. Cartesian gradient-echo
coronary magnetic resonance angiography. Magn
Reson Med 2001;46:789 94.
[179] Nishimura DG, Irarrazabal P, Meyer CH. A velocity
k-space analysis of flow effects in echo-planar and
spiral imaging. Magn Reson Med 1995;33:549 56.
[180] Deshpande VS, Shea SM, Chung YC, McCarthy RM,
Finn JP, Li D. Breath-hold three-dimensional trueFISP imaging of coronary arteries using asymmetric
sampling. J Magn Reson Imaging 2002;15:473 8.
[181] Yang PC, Meyer CH, Terashima M, et al. Spiral
magnetic resonance coronary angiography with rapid
real-time localization. J Am Coll Cardiol 2003;41:
1134 41.
[182] Larson AC, Simonetti OP, Li D. Coronary MRA with
3D undersampled projection reconstruction TrueFISP.
Magn Reson Med 2002;48:594 601.
[183] Azhari H, Sodickson DK, Edelman RR. Rapid MR
imaging by sensitivity profile indexing and deconvolution reconstruction (SPID). Magn Reson Imaging
2003;21:575 84.
[184] Sodickson DK, McKenzie CA, Ohliger MA, Yeh EN,
Price MD. Recent advances in image reconstruction,
coil sensitivity calibration, and coil array design for
SMASH and generalized parallel MRI. Magma 2002;
13:158 63.
[185] Sodickson DK, McKenzie CA. A generalized approach to parallel magnetic resonance imaging. Med
Phys 2001;28:1629 43.
[186] Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: sensitivity encoding for fast MRI.
Magn Reson Med 1999;42:952 62.
[187] Weiger M, Pruessmann KP, Boesiger P. Cardiac realtime imaging using SENSE. SENSitivity Encoding
scheme. Magn Reson Med 2000;43:177 84.
[188] Weiger M, Pruessmann KP, Kassner A, et al. Contrast-enhanced 3D MRA using SENSE. J Magn Reson
Imaging 2000;12:671 7.
[189] Stuber M, Botnar RM, Fischer SE, et al. Preliminary
report on in vivo coronary MRA at 3 Tesla in
humans. Magn Reson Med 2002;48:425 9.
[190] Edelman RR, Manning WJ, Burstein D, Paulin S.
Coronary arteries: breath-hold MR angiography. Radiology 1991;181:641 3.
[191] Manning WJ, Li W, Edelman RR. A preliminary report comparing magnetic resonance coronary angiography with conventional angiography. N Engl J Med
1993;328:828 32.
[192] Manning WJ, Li W, Boyle NG, Edelman RR. Fat-
672
[193]
[194]
[195]
[196]
[197]
[198]
[199]
[200]
[201]
[202]
[203]
[204]
[205]
[206]

suppressed breath-hold magnetic resonance coronary
angiography. Circulation 1993;87:94 104.
Pennell DJ, Bogren HG, Keegan J, Firmin DN, Underwood SR. Assessment of coronary artery stenosis by
magnetic resonance imaging. Heart 1996;75: 127 33.
Spuentrup E, Buecker A, Stuber M, et al. Navigatorgated coronary magnetic resonance angiography using steady-state-free-precession: comparison to standard T2-prepared gradient-echo and spiral imaging.
Invest Radiol 2003;38:263 8.
Deshpande VS, Shea SM, Laub G, Simonetti OP,
Finn JP, Li D. 3D magnetization-prepared true-FISP:
a new technique for imaging coronary arteries. Magn
Reson Med 2001;46:494 502.
Shea SM, Deshpande VS, Chung YC, Li D. Threedimensional true-FISP imaging of the coronary arteries: improved contrast with T2-preparation. J Magn
McCarthy RM, Shea SM, Deshpande VS, et al. Coronary MR angiography: true FISP imaging improved
by prolonging breath holds with preoxygenation in
healthy volunteers. Radiology 2003;227:283 8.
Deshpande VS, Shea SM, Li D. Artifact reduction in
true-FISP imaging of the coronary arteries by adjusting imaging frequency. Magn Reson Med 2003;49:
803 9.
Zheng J, Li D, Bae KT, Woodard P, Haacke EM.
Three-dimensional gadolinium-enhanced coronary
magnetic resonance angiography: initial experience.
J Cardiovasc Magn Reson 1999;1:33 41.
Stuber M, Botnar RM, Danias PG, et al. Contrast
agent-enhanced, free-breathing, three-dimensional
coronary magnetic resonance angiography. J Magn
van Rossum AC, Bedaux WL, Hofman MB. Morphologic and functional evaluation of coronary artery
bypass conduits. J Magn Reson Imaging 1999;10:
734 40.
Nwasokwa ON. Coronary artery bypass graft disease.
Ann Intern Med 1995;123:528 45.
Campeau L, Enjalbert M, Lesperance J, Vaislic C,
Grondin CM, Bourassa MG. Atherosclerosis and late
closure of aortocoronary saphenous vein grafts: sequential angiographic studies at 2 weeks, 1 year,
5 to 7 years, and 10 to 12 years after surgery. Circulation 1983;68(3 Pt 2):II1 7.
Bourassa MG, Fisher LD, Campeau L, Gillespie MJ,
McConney M, Lesperance J. Long-term fate of
bypass grafts: the Coronary Artery Surgery Study
(CASS) and Montreal Heart Institute experiences.
Circulation 1985;72(6 Pt 2):V71 8.
Botnar RM, Stuber M, Danias PG, Kissinger KV,
Manning WJ. Improved coronary artery definition
with T2-weighted, free-breathing, three-dimensional
coronary MRA. Circulation 1999;99:3139 48.
Stuber M, Botnar RM, Danias PG, et al. Doubleoblique free-breathing high resolution three-dimensional coronary magnetic resonance angiography.
[207] Rubinstein RI, Askenase AD, Thickman D, Feldman

MS, Agarwal JB, Helfant RH. Magnetic resonance
imaging to evaluate patency of aortocoronary bypass
grafts. Circulation 1987;76:786 91.
[208] White R, Caputo G, Mark A, Modin G, Higgins C.
Coronary artery bypass graft patency: noninvasive
evaluation with MR imaging. Radiology 1987;164:
681 6.
[209] Jenkins JP, Love HG, Foster CJ, Isherwood I, Rowlands DJ. Detection of coronary artery bypass graft
patency as assessed by magnetic resonance imaging.
Br J Radiol 1988;61:2 4.
[210] Aurigemma GP, Reichek N, Axel L, Schiebler M,
Harris C, Kressel HY. Noninvasive determination of
coronary artery bypass graft patency by cine magnetic
resonance imaging. Circulation 1989;80:1595 602.
[211] Galjee MA, van Rossum AC, Doesburg T, van Eenige
MJ, Visser CA. Value of magnetic resonance imaging
in assessing patency and function of coronary artery
bypass grafts: an angiographically controlled study.
[212] Brenner P, Wintersperger B, von Smekal A, et al.
Detection of coronary artery bypass graft patency
by contrast enhanced magnetic resonance angiography. Eur J Cardiothorac Surg 1999;15:389 93.
[213] Vrachliotis TG, Bis KG, Aliabadi D, Shetty AN,
Safian R, Simonetti O. Contrast-enhanced breathhold MR angiography for evaluating patency of
coronary artery bypass grafts. AJR Am J Roentgenol
1997;168:1073 80.
[214] Achenbach S, Kessler W, Moshage WE, et al. Visualization of the coronary arteries in three-dimensional
reconstructions using respiratory gated magnetic
resonance imaging. Coron Artery Dis 1997;8:441 8.
[215] Kessler W, Achenbach S, Moshage W, et al. Usefulness of respiratory gated magnetic resonance coronary angiography in assessing narrowings 50%
in diameter in native coronary arteries and in aortocoronary bypass conduits. Am J Cardiol 1997;80:
989 93.
[216] Langerak SE, Vliegen HW, de Roos A, et al. Detection of vein graft disease using high-resolution magnetic resonance angiography. Circulation 2002;105:
328 33.
[217] Langerak SE, Kunz P, Vliegen HW, et al. MR flow
mapping in coronary artery bypass grafts: a validation
study with Doppler flow measurements. Radiology
2002;222:127 35.
[218] Langerak SE, Vliegen HW, Jukema JW, et al. Value
of magnetic resonance imaging for the noninvasive
detection of stenosis in coronary artery bypass grafts
and recipient coronary arteries. Circulation 2003;107:
1502 8.
[219] Galjee MA, van Rossum AC, Doesburg T, Hofman
MB, Falke TH, Visser CA. Quantification of coronary
artery bypass graft flow by magnetic resonance phase
velocity mapping. Magn Reson Imaging 1996;14:
485 93.
[220] Hoogendoorn LI, Pattynama PM, Buis B, van der
[221]
[222]
[223]
[224]
[225]
Geest RJ, van der Wall EE, de Roos A. Noninvasive

evaluation of aortocoronary bypass grafts with magnetic resonance flow mapping. Am J Cardiol 1995;
75:845 8.
Langerak SE, Kunz P, Vliegen HW, et al. Improved
MR flow mapping in coronary artery bypass grafts
during adenosine-induced stress. Radiology 2001;
218:540 7.
Kim WY, Danias PG, Stuber M, et al. Coronary magnetic resonance angiography for the detection of
coronary stenoses. N Engl J Med 2001;345:1863 9.
Hejmadi A, Sahn DJ. What is the most effective method of detecting anomalous coronary origin in symptomatic patients? J Am Coll Cardiol 2003;42: 155 7.
Post JC, van Rossum AC, Bronzwaer JG, et al. Magnetic resonance angiography of anomalous coronary
arteries: a new gold standard for delineating the proximal course? Circulation 1995;92:3163 71.
Bunce NH, Rahman SL, Keegan J, Gatehouse PD,
Lorenz CH, Pennell DJ. Anomalous coronary arteries:
anatomic and functional assessment by coronary and
[226]
[227]
[228]
[229]
673
perfusion cardiovascular magnetic resonance in three

sisters. J Cardiovasc Magn Reson 2001;3:361 9.
McConnell MV, Stuber M, Manning WJ. Clinical role
of coronary magnetic resonance angiography in the
diagnosis of anomalous coronary arteries. J Cardiovasc Magn Reson 2000;2:217 24.
Wittlinger T, Voigtlander T, Rohr M, et al. Magnetic
resonance imaging of coronary artery occlusions in the
navigator technique. Int J Cardiovasc Imaging 2002;
18:203 11 [discussion: 213].
Watanabe Y, Nagayama M, Amoh Y, et al. High-resolution selective three-dimensional magnetic resonance coronary angiography with navigator-echo
technique: segment-by-segment evaluation of coronary artery stenosis. J Magn Reson Imaging 2002;
16:238 45.
Krittayaphong R, Mahanonda N, Kangkagate C,
Nakyen S, Tanapibunpon P, Chaithiraphan S. Accuracy of magnetic resonance imaging in the diagnosis
of coronary artery disease. J Med Assoc Thai 2003;
86(Suppl 1):S59 66.
Radiol Clin N Am 42 (2004) 675 690
Congenital heart disease in the adult: 2004

Murray G. Baron, MDa,*, Wendy M. Book, MDb
b
a
Department of Radiology, Emory University Hospital, 1364 Clifton Road Northeast, Atlanta, GA 30322, USA
Division of Cardiology, Department of Internal Medicine, Emory University Hospital, 1364 Clifton Road Northeast,
Atlanta, GA 30322, USA
Until recently, the population of adults with congenital heart disease has been small, with little impact
on the overall practice of radiology. The number of
patients remained stable over the years because the
birth rate of children with cardiac anomalies varies
little and, in the absence of effective means for
treatment, so does the annual death rate. This has
changed radically in the last five or so decades
because of advances in open heart surgery, anesthesiology, and cardiac imaging techniques. Most conditions that used to be almost automatically fatal in
the first years of life can now be corrected or successfully palliated so that, at present, in this country
about 85% of all infants with congenital heart disease
survive into adult life [1]. Using statistics from the
mid-1980s forward, it has been estimated that in the
year 2000 there were about 787,000 patients with
congenital heart disease of all degrees of severity in
the United States, treated and untreated, but excluding the numerous cases of isolated bicuspid aortic
valve [2]. This number increases by about 5% each
year and almost certainly exceeds 1 million as of
2004 and continues to enlarge.
Although correction of a lesion may re-establish a
relatively normal pattern of blood flow, more than
half of adult survivors are at significant risk for
developing complications either from their operative
procedures or from lingering effects of the original
lesion. Many of these complications can be managed
effectively or corrected if detected promptly. Patients
with congenital heart disease, regardless of their stage
of treatment, should be followed with periodic examinations. Because most of these examinations and
E-mail address: mbaron@emory.edu (M.G. Baron).
those related to episodes of intercurrent disease include radiograph studies, it is likely that most radiologists at one time or another are involved with the
care of congenital cardiac patients.
Bicuspid aortic valve

The bicuspid aortic valve is the most common
cardiac anomaly seen in adults [3,4], occurring in
about 2% of all births [5]. The valve is functionally
competent in children but is prone to accelerated
degeneration because the leaflets are poorly fitted to
the valve orifice and usually have broad, redundant
areas of contact that slap as they close, compared with
the smooth diastolic coaptation of the leaflets of a
normal, tricuspid aortic valve. Fibrosis and dystrophic
calcification of the leaflets is common, often leading
to stenosis. So long as the valve orifice is not
significantly narrowed, the appearance of the heart
on chest films is normal. When the valve is stenotic,
the cardiac silhouette has the same appearance as with
rheumatic aortic stenosis or degenerative stenosis
associated with aging, except that calcification of a
bicuspid valve tends to be more extensive and occurs
earlier, frequently before the age of 40. Poststenotic
dilatation of the aorta is seen in some cases of bicuspid
valve stenosis.
Atrial septal defect

The salient plain film findings are cardiomegaly,
enlargement of the right ventricle, dilation of the
main pulmonary artery, and pulmonary plethora
(Fig. 1) [3,4]. Unlike children with an atrial septal
defect, in whom left atrial enlargement is rare, almost
doi:10.1016/j.rcl.2004.03.008
676
M.G. Baron, W.M. Book / Radiol Clin N Am 42 (2004) 675690
Fig. 1. Atrial septal defect, woman in early 40s. The heart is

slightly enlarged, the main pulmonary artery (arrow) is
markedly dilated, and the peripheral pulmonary vasculature
is increased. The aorta is characteristically small.
half of adults over the age of 40 with an atrial septal

defect have dilatation of this chamber [6].
Most patients with an atrial septal defect have
some degree of pulmonary hypertension because of
the increased flow of blood through the pulmonary
circulation, but only about 10% of adults, more
commonly women, develop resistive pulmonary hypertension (Eisenmengers syndrome) [7]. Instead of
the generalized increase in caliber of the pulmonary
arteries and veins associated with flow hypertension,
only the main pulmonary artery and hilar arteries are
dilated when the hypertension is caused by increased
arteriolar resistance. In these cases, the peripheral
arteries are diminished in size, creating the appearance of pruning of these vessels, whereas the
pulmonary veins remain within normal limits in size
(Fig. 2). These findings are similar to those of primary
pulmonary hypertension, although the clinical course
of the Eisenmengers syndrome is considerably more
benign than the relatively rapid, fatal progression of
primary pulmonary hypertension [8]. Mural calcification of the central pulmonary arteries can occur with
either type of pulmonary hypertension, probably
resulting from the prolonged overdistention of the
vessels and damage to their walls. Once the atrial
septal defect is closed, the pulmonary plethora recedes
but the changes of resistive hypertension usually
remain unchanged and, in an occasional instance,
may even progress after closure of the defect.
Recently, percutaneous catheter-delivered devices
have become popular for closure of atrial septal
defects in children and adults. These are usually
composed of two umbrella-like or clam-shell segments that clamp the atrial septum between them and
cover the defect. The metallic center posts and struts,
if present, of the devices can be seen in the central

portion of the cardiac silhouette on a frontal film and
posterior to the midline in the lateral view (Fig. 3).
The devices are not used to close sinus venosus
defects, which are in the upper atrial septum, because
of the proximity of the superior vena caval orifice,
and are not seen in the superior portion of the septum.
The atrial defect occluders should not be confused
with the PLAATO device, which is used to occlude
the left atrial appendage. These are seen near the left
border of the cardiac silhouette in the frontal view,
just below the pulmonary artery segment, and slightly
more posterior than the atrial septal defect device in
the lateral view (Fig. 4).
Pulmonic valve stenosis

The radiographic picture of mild and moderate
pulmonic valve stenosis [3,4] consists of a normalsized heart, often with rounding of the cardiac apex
because of right ventricular hypertrophy; poststenotic
dilatation of the main pulmonary artery; and dilatation and lateral displacement of the left pulmonary
artery (Fig. 5). With more severe disease the right
ventricle is often dilated and right atrial enlargement
can occur, particularly if there is tricuspid insuffi-
Fig. 2. Atrial septal defect with Eisenmengers syndrome in

a 60-year-old man. The heart is moderately enlarged with
globular dilatation of the central pulmonary arteries. Beyond
the hila, there is a sudden decrease in the caliber of the
pulmonary vessels (pruning) indicative of resistive pulmonary hypertension. Right ventricular pressure, 91/7; femoral
artery oxygen saturation, 89%.
677
Fig. 3. Transcatheter occlusion of atrial septal defect in a 42-year-old woman with a Cardioseal device (arrows). (A) Frontal view.
(B) Lateral view.
ciency. The pulmonary vascularity in uncomplicated

pulmonic valve stenosis is normal.
Mild and moderate degrees of stenosis are most
commonly treated with balloon valvotomy. The radiographic appearance of the heart does not change significantly when correction is performed in adult life.
Coarctation of the aorta

Coarctation of the aorta [3,4] is a potentially
lethal lesion. If not treated, it is estimated that 60%
of affected infants die in the first year of life and
about 25% of the remainder are dead by the age of
Fig. 4. PLAATO (arrows) device delivered by catheter into the left atrial appendage to prevent dissemination of possible future
thrombi. Patient has atrial fibrillation. (A) Frontal projection. (B) Lateral projection.
678
Fig. 5. Pulmonic valve stenosis. Normal-sized heart, dilated

main pulmonary artery, and prominent left pulmonary artery.
The course of this vessel is posterolateral rather than the
almost directly posterior course seen in the normal patient.
in conjunction with notching of the ribs (Fig. 8), but

rib notching by itself is not diagnostic of coarctation
and also occurs in neurofibromatosis or arteriovenous
malformations of an upper extremity. Conversely,
absence of rib notching does not exclude aortic
coarctation because it may not be detected in as many
as 25% of adults with a clinically significant lesion.
Although 80% of patients whose coarctation
was corrected in childhood are still alive 40 years
postsurgery, they still do not have a normal life expectancy because of recoarctation, persistence or
recurrence of systemic arterial hypertension, accelerated atherosclerosis and coronary artery disease, or
rupture of an aortic aneurysm.
Unlike the original coarctation, recoarctation is
unlikely to be detected on routine chest films because
the characteristic contours of the abnormal region of
the aorta are largely obscured by changes from the
prior surgery. The recurrent narrowing of the aortic
lumen can be delineated accurately by MR imaging
or contrast-enhanced multislice CT studies. Persistent
hypertension and atherosclerosis produce plain film
images no different from those seen with either of
20 years [9]. Adults in whom the disease was not

detected earlier usually come to medical attention
because of complications of arterial hypertension in
the upper body or, on occasion, because of an abnormal chest radiograph.
Most coarctations occur in the proximal descending aorta, at the level of the ligamentum arteriosum.
The left subclavian artery usually arises from the
aorta immediately above the coarcted segment and
the root of the vessel is dilated forming a bulge just
above the coarctation, overlying and obscuring the
aortic knob. If there is poststenotic dilatation of the
aorta, the additional bulge below the coarcted segment
gives the upper left mediastinal contour a figure-ofthree configuration, an appearance characteristic of
coarctation. This is not commonly seen in the adult.
More often, the aortic knob appears abnormal
because it does not have an upper margin. What is
seen as a bulge just above the coarctation is not the
aortic knob, but the dilated root of the left subclavian
artery and its upward and outward curve as it extends
toward the left arm (Fig. 6). The upper margin of a
normal aortic knob is outlined by air-containing lung
and forms a horizontally or obliquely oriented arc that
curves into the mediastinum (Fig. 7). Absence of the
roof of the aortic knob is the most common sign of
coarctation in the adult, but is not pathognomonic
(eg, it also can be caused by enlarged left superior
mediastinal lymph nodes or postpleuritic scarring).
This abnormality of the knob is diagnostic when seen
Fig. 6. Coarctation of the aorta. Selective aortogram. The left

subclavian artery (LS) arises from the descending aorta just
above the coarctation (white arrow) and extends upward and
outward toward the left arm. Its root is dilated and overlies
and obscures the aortic knob (black arrows). AA, ascending
aorta; In, innominate artery.
679
Fig. 7. The aortic knob. (A) Normal aortic knob. The aortic knob is formed by the most distal portion of the aortic arch, as the
vessel curves downward to become the descending aorta. Its upper margin (arrow) is normally clearly silhouetted against the
adjacent air-containing lung. (B) In coarctation, the aortic knob is obscured by the left subclavian artery and the expected incisura
between the aortic knob and mediastinum is not visible.
these conditions in the absence of coarctation except

that the degenerative changes tend to appear at a
younger age. The average age of acute myocardial
infarction in patients with repaired coarctation is
53 years [10,11]. The authors know of no specific
coronary calcium screening study following coarctation repair, but it seems reasonable to expect higher
calcium scores at younger ages in these patients [12].
About 9% of patients develop aortic aneurysms as
a late complication of surgical repair [13], first
appearing 10 to 15 years after the correction. Most
Fig. 8. Coarctation of the aorta. Absence of a normal aortic

knob with smoothing of the mediastinal contour (arrowhead) together with rib notching (small arrows) is diagnostic
of coarctation of the aorta.
often the aneurysm occurs in the region of the

previous coarctation but involvement of the ascending aorta is not rare.
There are no differences in the postoperative
blood pressure levels or other clinical findings between those patients who do and those who do not
develop an aneurysm. The only statistically significant predictors are the age of the patient at the time of
correction (a lower incidence when surgery was
performed in the early years of life) and the use of
a patch graft, particularly Dacron, to widen the
coarcted segment [13]. As a result, the technique
has been largely abandoned but a fair number of
such patients are still living and remain at risk for
aneurysm formation.
Aneurysms that form in patients with patch grafts
usually occur on the greater curvature of the aorta and
include the grafted segment. On occasion, however,
the aneurysm under similar circumstances involves
the aortic wall opposite the patch, on the lesser curve
of the aorta. Growth rate of an aneurysm is quite
variable and on occasion, they suddenly enlarge quite
rapidly. As a rule, aneurysms tend to be asymptomatic until shortly before they rupture. The risk of
rupture is particularly high in women during pregnancy. Elective repair of an aneurysm does not entail
an excessive operative risk but the mortality rate
increases sharply once rupture occurs.
The plain chest film is a reasonable screening
method for aneurysms of the descending aorta
(Fig. 9), having a high sensitivity but low specificity
680
Fig. 9. Coarctation of the aorta. Aneurysm at site of repair. (A) Frontal view. The aneurysm (arrow) does not obliterate the
smoothened border of the mediastinum because it is situated posteriorly on the descending aorta. (B) Lateral view.
[14]. Almost all aneurysms at the coarctation site are

detected but the false-positive rate is unacceptably
high and positive findings should be confirmed by
other, more accurate, imaging methods that can
demonstrate the true size of the aortic lumen.
Following surgery, slight to moderate dilatation of
the aortic lumen is common at the site of the previous
coarctation and should not be confused with an
aneurysm. Using absolute limits of the aortic diameter to define an aneurysm, as is done with atherosclerotic aneurysms in the adult, is not satisfactory for
these patients because a significant number are still in
their growing years. It is more accurate to normalize
the differences caused by growth by comparing the
aortic diameter at the repair site with its diameter at
the level of the diaphragm. A ratio of 1.5 or greater is
considered indicative of an aneurysm [15]. This value
was originally determined from measurements on
aortograms but multislice CT or MR images should
produce equivalent results. A caveat for the use
of tomographic techniques is that the diameter is
overestimated unless it is measured from a true
cross-sectional view of the aorta. On MR imaging
examinations the desired plane of acquisition through
the descending aorta can be plotted from sagittal and
coronal scout views, whereas with thin-section CT
postprocessing with three-dimensional reformatting
of the original data produces comparable results.
Moderate dilatation of the aortic root is not uncommon in adults following coarctation repair. At
times, the sinuses of Valsalva also are enlarged,
resulting in annuloaortic ectasia. In one series of

92 patients in whom the aortic root was evaluated,
dilatation was present in 24 (26%) [11]. Aneurysms of
the ascending aorta, which may be saccular or dissecting, are less common than of the descending aorta
but are of greater clinical significance because of the
risk of rupture. The only statistically significant finding of predictive value for the development of such
aneurysms, by multivariate analysis, is the presence of
a bicuspid aortic valve. Although such a valve is
common in patients with coarctation (the incidence
varies wildly in published series because of patient
selection, but is likely about 40% or slightly higher), it
is much more common in coarctation patients with an
ascending aortic aneurysm, often reaching 80% or
more in published cases [11,13]. This is not surprising
considering that both light and electron microscopic
examination of biopsies of the ascending aortic wall
from patients with bicuspid valves show degenerative
changes in the media similar to those seen with
Marfan syndrome, regardless of whether the valve is
stenotic or insufficient, or is functioning normally
[16]. Similar aneurysms have been reported with a
bicuspid valve in the absence of coarctation.
Aneurysms of the ascending aorta are often difficult to see on routine frontal and lateral chest films.
To improve the sensitivity, a left anterior oblique
view, which shows the anterior margin of the ascending aorta in profile, away from the sternum, is
suggested as part of the standard radiographic examination when evaluating postcoarctation patients.
Tetralogy of Fallot
At an early stage in development, before the heart
divides into four separate chambers, the two ventricles have a single, common outflow tract, the
conus, which leads to a common outflow vessel,
the truncus arteriosus. As the embryo grows, the
conus is divided by a midline septum, which joins
the interventricular septum so that the right ventricle
communicates with the more anterior conal channel
and the left ventricle with the posterior one. At about
the same time, a septum forms in the midline of the
truncus and joins the conus septum. The right ventricular conus channel communicates with the anterior
truncal channel, which develops into the main pulmonary artery, and the left ventricular conus channel
becomes continuous with the posterior truncal compartment, the future systemic aorta [17].
If the conus septum develops anterior to the
midline of the conus the outflow tract of the right
ventricle is narrowed, appearing in the mature heart
as infundibular stenosis, often with stenosis or hypoplasia of the pulmonary valve. In the most severe
case, the pulmonic valve is atretic. Because the
conus septum is in the wrong position, its lower
end is misaligned with the interventricular septum
and the two do not meet, leaving a large ventricular
septal defect through which the two ventricles intercommunicate freely. The aortic root is usually enlarged, extending anteriorly to override the septal
defect so that it receives blood from both ventricles.
As a result of the right-to-left shunt, unoxygenated
blood is directly recirculated to the body and the
patient is cyanotic.
Almost all tetrads are detected early in life. Of the
few who slip into adulthood untouched, most have
minimal forms of the anomaly and slight, if any,
cyanosis (pink tetrads). These patients usually have
normal-appearing chest films, except for the common
association of a mirror image right aortic arch, a
finding common to all forms of tetralogy.
Before the advent of effective cardiopulmonary
bypass, the surgical approach to tetralogy was limited
to creation of a palliative, extracardiac, left-to-right
shunt, designed to bring more of the poorly saturated
systemic blood to the pulmonary circulation and so
improve overall oxygenation. The common anastomoses were (1) the Blalock-Taussig shunt, in which a
subclavian artery was transected and its proximal
portion connected end-to-side to a pulmonary artery;
(2) the Waterston shunt, a side-to-side anastomosis
between the ascending aorta and the right pulmonary
artery; and (3) the Potts shunt, a side-to-side connection between the descending aorta and the left main
681
pulmonary artery. Major problems shared by these

procedures are the difficulty in gauging the correct
size of the shunt orifice so that the volume of blood
flowing to the pulmonary circuit is adequate but not
excessive, and avoidance of traction at the anastomotic site to prevent kinking of one or both pulmonary arteries.
Some adults are still encountered with well-functioning shunts as their only surgical treatment. The
pulmonary vasculature may appear normal unless one
pulmonary artery is kinked, causing a disparity between the vascularity of the two lungs (Fig. 10). An
overly large shunt causes pulmonary plethora, similar
to any other left-to-right shunt, and if it eventuates in
resistive pulmonary hypertension, dilatation of the
central pulmonary arteries and pruning of the peripheral vessels. The cardiac silhouette still tends to be
boot-shaped, the toe of the boot formed by the
blunted cardiac apex because of right ventricular
hypertrophy and the concave instep reflecting the
small infundibulum and pulmonary artery. This appearance is most marked in cases with pulmonary
atresia (Fig. 11).
The first intracardiac repair of a tetralogy was
achieved in 1954 [18]. Total correction involves
closure of the ventricular septal defect and establish-
Fig. 10. Boot-shaped heart in a 43-year-old woman with

pulmonary atresia and ventricular septal defect. The heart is
slightly enlarged with an elevated, blunted apex caused by
right ventricular hypertrophy and dilatation. The normal
bulge of the right ventricular outflow tract and main
pulmonary artery below the aortic knob segment is absent
because these structures are hypoplastic. This creates the
concave instep of the boot (arrow). The pulmonary
vasculature is disorganized because it is supplied entirely
by bronchial artery collaterals.
682
Fig. 11. Left pulmonary artery stenosis at site of Potts shunt.

Gadolinium angiogram, left oblique view. The stenosis of the
main pulmonary artery (L) is at the site of a Potts shunt,
where the vessel approximates the descending aorta (D).
Marked poststenotic dilatation is seen beyond the narrowing
of the vessel. AA, ascending aorta; PA, main pulmonary
artery; RV, right ventricle.
ment of a nonrestrictive outflow pathway for the right

ventricle. In some cases, the latter only requires
resection of infundibular musculature and possibly
pulmonary valvotomy. If a significant pressure gradient is still present because the infundibular channel
is too narrow, a lengthwise incision is made in the
anterior wall of the infundibulum, the edges spread
apart and roofed over with a pericardial or synthetic
patch. If the annulus of the pulmonic valve also is
hypoplastic, the incision can be extended onto the
main pulmonary artery and a larger, transannular patch
used (Fig. 12).
If the pulmonic valve is atretic, the ventricular
septal defect is closed and the valve region bypassed
by a conduit extending from the right ventricle to a
pulmonary artery, which can include a tissue valve.
These conduits may have to be replaced several times
during the life of the patient because of degenerative
changes, usually formation of an intimal peel that
narrows the lumen or calcification of the conduit wall
and valve, or because the fixed size of the conduit
becomes inadequate as the patient grows.
Twenty-year survival rates of 90% or better are
common for complete correction of tetralogy of Fallot;
however, this does not signify a return to normalcy.

Late complications occur in a significant number of
patients, most often involving pulmonary valve insufficiency and right ventricular dysfunction and these
tend to increase in frequency as the postoperative
interval lengthens [19].
Some degree of insufficiency of the pulmonic
valve is present in almost all patients following
correction of tetralogy. The actual incidence and
severity is partly dependent on the extent of the
corrective procedure, the greatest occurring when
the pulmonary valve annulus is violated by a transannular patch, but it is also common in patients
whose patch is limited just to the right ventricular
outflow tract [20]. Simply performing a pulmonary
valvotomy, even in the absence of tetralogy, usually
results in some degree of valvular insufficiency.
At one time it was thought that the heart could
tolerate relatively severe degrees of pulmonic insufficiency with impunity. This has been shown to be
incorrect. Although almost 90% of corrected tetrads
are asymptomatic and their right ventricles show
normal functional values at rest, when stressed these
values tend to be significantly lower than normal
[21,22].
The chronic regurgitation of blood through the
pulmonic valve results in volume overload of the
right ventricle, resulting in dilatation of the chamber
and myocardial hypertrophy. At first the hypertrophy
is adequate to maintain normal stroke volume and the
ejection fraction remains within normal limits. With
time, as the chamber dilates further ventricular function becomes impaired, relative insufficiency of the
tricuspid valve may develop, and the ejection fraction
drops. At this point, cardiac enlargement is usually
obvious on plain films.
In the past 10 years or so, stenosis of the pulmonic
valve has been treated in many cases by percutaneous
balloon valvotomy. Although this produces only
minor degrees of valve insufficiency, experience with
the technique is too short to predict whether this will
remain inconsequential.
The effects of pulmonary insufficiency on the
right ventricle are accentuated if there is a significant
postoperative systolic gradient because of residual
narrowing of the ventricular outflow tract or pulmonary artery. In some instances, this may be caused by
stenosis of a pulmonary artery secondary to closure of
an earlier systemic-to-pulmonary shunt. Persistence
of a ventricular shunt because of incomplete closure
of the ventricular septal defect, dehiscence of the
surgical patch, or undetected muscular ventricular
septal defects also accentuate right ventricular dysfunction, as does tricuspid valve insufficiency.
683
Fig. 12. Tetralogy with hypoplastic right ventricular outflow tract and pulmonic valvular stenosis. (A) Selective right
ventriculogram: systole. The entire infundibulum (In) is narrowed, as are the pulmonic valve annulus and main pulmonary artery.
Bowing of the pulmonic valve away from the ventricle indicates pulmonic valve stenosis (arrowheads). A transannular patch
was used when this patient was repaired. The left pulmonary artery (LPA) shows poststenotic dilatation and follows an abnormal
posterolateral course rather than going directly posterior as in the normal. The ascending aorta (Ao) is faintly opacified because
of the right-to-left shunt through the ventricular septal defect (not seen in this section). (B) Coronal section, spin echo image of
another patient, postcorrection of tetralogy with a transannular patch. The infundibulum (In) is wide open and the patch bulges
outward along the left side of the heart. RA, right atrium; RV, right ventricle.
Although chronic, severe pulmonary insufficiency

by itself can result in right ventricular dysfunction,
other factors are also present in postoperative tetrads.
Probably the most important of these is the infundibular incision and resection of infundibular muscle.
There is evidence suggesting that normally functioning infundibular musculature tends to protect the right
ventricle from the effects of pulmonary insufficiency
[23]. The ventricular outflow tract can also affect
right ventricular function in other ways. Early surgical corrections often used pliable infundibular or
transannular patches to ensure a wide-open outflow
path for the right ventricle. Occasionally, such a patch
acts like a ventricular aneurysm, distending as the
ventricle contracts and acting as a sump, diminishing
the effectiveness of ventricular systole [21]. The
authors know of no reported case where such a patch
has ruptured.
Bulging of the patch produces a localized protrusion from the left cardiac contour in the region
normally formed by the left atrial appendage. Distinction between bulging of a ventricular patch and a
dilated appendage can be difficult, especially when
the bulge is small, unless there is an associated right
aortic arch to suggest the possibility of a tetralogy
(Fig. 13A). If the bulge is larger and extends onto the
area of the main pulmonary artery, an enlarged left
atrial appendage becomes considerably less likely
(Fig. 13B). Identification of a bulging patch is not

an indication for surgery, but the area should be
watched on follow-up examinations for progressive enlargement.
The presence of a prosthetic valve in the pulmonic position (at least in the lateral view) is also
suggestive of a postoperative tetralogy (Fig. 14). A
prosthetic valve in a right ventricle-to-pulmonary
artery conduit is often seen further to the right than
the native pulmonary valve in the frontal view, but
still abuts the anterior aspect of the heart in the
lateral. Replacement of this valve is very rare in
cases of isolated pulmonic stenosis. Acquired diseases producing significant distortion or destruction
of the valve are limited to bacterial endocarditis
(usually in drug addicts) and cases of the carcinoid syndrome.
Even without a patch, scarring from a longitudinal incision in the wall of the outflow tract interferes
with right ventricular function because the scarring
impedes normal infundibular contraction in late systole, and decreases the ejectile thrust of the ventricle [23].
Recent MR imaging studies of corrected tetrads
have shown impairment of left and right ventricular
function in some patients. Increase in left ventricular
end-diastolic volume and decrease in its output are directly related to the degree of pulmonic insufficiency
684
Fig. 13. Status postrepair of tetralogy of Fallot. (A) An infundibular patch (arrow) bulges from the left border of the cardiac
silhouette in the same region as does a dilated left atrial appendage. This patient has a right aortic arch (A), however, which
should suggest the possibility of a tetralogy. (B) This patient required a transannular patch and the bulge is considerably more
extensive than in the previous case. The patch (arrow) extends almost to the roof of the main pulmonary artery, distinguishing
it from the more limited bulge of a dilated left atrial appendage. Ao, aortic knob.
and the severity of right ventricular dysfunction

[24,25].
The diameter of the ascending aorta is often larger
than normal in tetrads, its size tending to be inversely
related to the degree of right ventricular outflow tract
hypoplasia. This is partly caused by the increased

volume of aortic blood flow and in part by the degree
of anterior displacement of the conus septum. In a
small subset of postoperative tetrads amounting to
15% of one series of 236 patients [26], the ascending
Fig. 14. Tetralogy of Fallot. Because of the degree of hypoplasia of the pulmonary artery, a valved conduit extending from the
right ventricle to the pulmonary artery was inserted at age 6 and replaced once thereafter, about 22 years ago. (A) Frontal view.
The valve ring is too high for an aortic valve and is somewhat more medial than usual for a pulmonic valve. (B) Lateral view.
The valve lies far anterior in the heart as does a normal pulmonic valve.
aorta was significantly larger than in the rest. Affected

patients were selected by comparing their aortic root
diameter with values from a nomogram of normal
aortic root diameters, normalized for body surface
area and age [27]. Significant aortic root dilatation
was defined as a ratio of observed to expected aortic
diameter exceeding 1.5. Progressive enlargement
with time is much more common in the dilated group
than in other tetrads.
The incidence of aortic root dilatation is greater in
patients with pulmonary atresia and ventricular septal
defect than in other tetrads and the size of the aorta
may continue to enlarge even after the anomaly has
been corrected. Eventually the root reaches a diameter where the aortic cusps can no longer coapt during
diastole and the resulting valvular insufficiency often
requires corrective surgery [28]. Dissecting aneurysm
has not been reported as a complication.
Transposition of the great vessels

Transposition of the great arteries results from
faulty division of the truncus so that the right ventricular outflow tract (derived from the anterior conal
channel) communicates with the aorta, whereas the
left ventricle ejects its blood through the main pulmonary artery to the lungs. This single malformation
results in two separate circulations, one from the right
ventricle through the aorta to the body and back to the
right heart, and the other from the left heart to the
lungs and back to the left heart, a situation that results
in the death of about 90% of untreated infants in the
first year of life [29]. The prognosis improves, at least
temporarily, if there is a second lesion, such as a
ventricular septal defect, that allows admixture of
blood from the two circulations. This formed the
basis for a reasonably successful early surgical approach in which the atrial septum was excised forming a single, common atrium where the oxygenated
and unoxygenated blood were mixed.
The primary goal of surgical therapy for complete
transposition is to redirect venous blood returning
from the body to the lungs so that it is oxygenated
before being pumped back to the body [30]. This was
first accomplished by switching blood flow at the
atrial level. The Senning operation (1959) formed a
conduit from the right atrial wall and the interatrial
septum that directed superior and inferior vena caval
blood to the mitral valve and the left atrium. The
Mustard operation (1963) accomplished the same
thing by excising the interatrial septum and constructing an internal baffle from pericardium or Dacron to
direct the caval flow to the mitral valve. This blue
685
blood is then pumped by the left ventricle to the

lungs where it is oxygenated and returned to the heart
by the pulmonary veins, then around the baffle and
through the tricuspid valve to the right ventricle and
then the aorta.
Either of these procedures, often performed in the
first year of life, had a relatively low operative
mortality in uncomplicated cases and a 10-year survival of 85% to 90%, and about 80% at 20 years
[29,31]. There is an increasing attrition rate, however,
in the later postoperative years. Even in the absence
of associated lesions, hemodynamic complications of
the atrial switch procedures appear in mid and late
survivors because of stenosis of venous inflow or
right ventricular dysfunction.
In 1975, Jatene et al [32] demonstrated the feasibility of correcting complete transposition by switching the aortic and main pulmonary artery trunks to
redirect the ventricular outflow. Systemic venous
blood flows through the right cardiac chambers and
is pumped to the lungs where it is oxygenated and
returned to the left side of the heart and aorta and then
to the systemic arterial circulation. Not only does this
reestablish a normal circulatory pathway, but each
ventricle is now returned to its normal physiologic
function within that circulation. Shortly thereafter,
atrial switch operations were largely abandoned. A
significant number of patients surviving an atrial
switch are now in their mid-adult years, however,
and late complications of those operations are still
being encountered. The older survivors of the arterial
switch operation are mostly in their twenties or
younger, and it will take another 10 or more years
before an accurate picture of the nature and incidence
of possible late complications becomes manifest.
Switch of the atrial inflow

With the exception of cardiac arrhythmias, the
common complications of switching the atrial inflow
are right ventricular dysfunction or stenosis of the
superior or inferior vena cava or the pulmonary veins.
The cause of the right ventricular dysfunction is
uncertain. There is no convincing evidence that the
intrinsic structure of the right ventricle is the culprit.
Although the two ventricles differ in shape and
myocardial fiber pattern, evidence from patients with
a different anomaly, congenitally corrected transposition, where the course of blood flow is normal but
the right ventricle must pump into the systemic
circuit, suggests that this chamber can tolerate the
chronic need for an almost fourfold increase in its
normal systolic pressure [33]. The chamber does
686
dilate moderately and its walls hypertrophy but its

function at rest tends to be adequate. A diminished
response to stress is common, but rarely severe.
Only 5% to 10% of patients following an atrial
switch procedure develop serious right ventricular
failure with progressive dilatation of the ventricle
leading to relative tricuspid insufficiency. At this
point, chest films show marked dilatation of the right
heart chambers.
The plain film appearance of the heart post atrial
switch is generally nonspecific. Most infants with
transposition of the great vessels show an oval
cardiac silhouette with a narrow vascular pedicle
(the egg-on-a-string appearance) because the two
great vessels are in line, the pulmonary artery situated
immediately behind the aorta. In about 15% of cases,
the vascular pedicle appears normal because the posterior pulmonary artery lies to one side or the other of
the aorta. Although the relationship of the great
vessels is not affected by the atrial switch operation,
the great vessels do enlarge with age and narrowing
of the pedicle is usually not obvious in the adult
(Fig. 15). This is particularly true if there is partial
obstruction of the superior vena caval orifice. The
cava then dilates and widens the superior mediastinum (Fig. 16).
Caval or pulmonary venous stenoses are direct
complications of the operative procedures, which
Fig. 15. Transposition of the great arteries in a 19-year-old

woman status postinteratrial baffle procedure. Heart is
moderately enlarged and nonspecific in shape. The vascular
pedicle is within normal limits. The aortic arch is on the left
(trachea displaced to the right) and the aortic knob is not
well seen.
Fig. 16. Transposition of the great arteries with superior

caval-atrial stenosis. Dilatation of the superior vena cava
widens the superior mediastinum to the right.
involve considerable suturing near the venous orifices, although problems frequently do not appear
until years after the surgery. Obstruction of pulmonary venous inflow is usually caused by development
of a stenotic ring within the left atrium, which affects
all pulmonary veins and results in the typical radiographic picture of congestive failure with pulmonary
venous distention and pulmonary edema. If only the
orifices of the right pulmonary veins, or less commonly the left veins, are compromised, the congestive
changes are limited to the involved lung.
Transposition of the great arteries, either untreated
or following an atrial switch, is usually obvious on
contrast-enhanced CT studies of the chest, even
without cardiac gating. If the aorta is positioned in
front of the pulmonary artery and the ventricles are in
normal position, the picture is diagnostic. This must
be distinguished from conditions in which the aorta is
so dilated that it extends further anteriorly than the
main pulmonary artery, as in some cases of tetralogy
of Fallot. In such instances, the aorta is mostly
alongside, and its posterior wall may still be behind,
the pulmonary artery. This is not a transposition.
With a well-functioning atrial switch procedure,
the abnormal circulatory pathway through the heart
and lungs may not be obvious on nongated scans
because both sides of the heart are usually opacified
when the images are obtained. When the superior
vena caval orifice is stenotic, however, the appearance is often specific. Because the obstruction occurs
at the cavoatrial junction, the higher pressure in the
proximal cava causes it to empty into the azygos
vein and reverse its flow (Fig. 17). Although some
flow into the azygos is not uncommon with normal
scans because of the pressure of injection or the
patient performing a Valsalvas maneuver, the retrograde flow is transitory, the azygos is of normal
687
caliber, and most of the contrast still fills the lower

superior cava as it flows into the right atrium. When
the cavoatrial junction is obstructed, the azygos vein
is dilated because of the increased volume of retrograde blood flow, and is densely and persistently
opacified. Both azygos and hemiazygos veins are
usually opacified, and the inferior vena cava because
of the normal communications between the two
venous systems.
Fig. 17. Transposition of the great arteries post Mustard operation with superior caval stenosis. (A) Axial section just below the
azygous-caval junction. The azygous vein (Az) is densely opacified because of retrograde flow from the cava (S). A, ascending
aorta in the position usually occupied by the pulmonary artery; P, pulmonary artery, posterior to the aorta; D, descending aorta.
(B) Section at level of bronchial bifurcation. The aortic valve (AV) lies anterior and slightly to the left of the main pulmonary
artery (P). Note the size of the azygous vein (Az), which is almost as large as the descending aorta (D). L, left pulmonary artery;
R, right pulmonary artery; S, superior vena cava. (C) Section at level of pulmonary veins. The right inferior pulmonary vein (RL)
and the left inferior vein (LL) enter the right atrium (RA) and are directed by the baffle through the mitral valve into the left
ventricle (LV). (D) Level of the ventricles. The inferior vena caval flow (IVC) flows through the mitral valve into the left
ventricle (LV).
688
Fig. 18. Anomalous origin of the right coronary artery from the left coronary sinus, 0.6-mm thick multislice CT sections. (A) The
right coronary artery is seen originating from the left coronary cusp, almost adjacent to the left coronary artery (not seen on this
section). It courses anteriorly between the right ventricular outflow tract (P) and the aorta (Ao). (B) A section, 1.2 mm more
caudad, shows the right coronary artery (arrow) turning downward into the atrial ventricular sulcus.
When the obstruction involves the inferior vena

caval orifice, it is usually not seen on chest CT scans
because the films are exposed before contrast material
returns from the peripheral circulation. If the contrast
was injected into a leg vein, it is likely that much of
the flow toward the heart is through the ascending
lumbar and azygos veins, similar to that seen with
interruption of the inferior vena cava.
Switch of the great arteries

The great arterial trunks are severed just above the
sinuses of Valsalva and switched so that the aorta
arises from the left ventricle and the pulmonary artery
from the right. The coronary arteries arise from the
sinuses of Valsalva above the right ventricle, however, and must be reimplanted above the left ventricle
after the aortic trunk has been switched. Coronary
artery stenoses and occlusions leading to left ventricular dysfunction are complications of these operations but usually become manifest in the immediate
postoperative and perioperative periods. Follow-up
studies at 5 and 10 years have not revealed additional
complications other than supravalvular pulmonic stenosis and insufficiency of the neoaortic left ventricular valve [34]. Whether or not the incidence of
intrinsic coronary disease increases more rapidly with
advancing age, or if stenosis caused by kinking of the
vessels occurs as the heart shape adapts to its changing hemodynamics and growth is not known. To date,
no serial studies have been published with CT or
MR imaging visualization of the coronary arteries in
these patients.
Anomalous aortic connection of a coronary artery

Origin of the left coronary artery from the right
coronary sinus, by itself or together with the right
coronary artery, or origin of both vessels from the
left coronary sinus are rare lesions. Usually, the
anomalous vessel courses between the aorta and
pulmonary artery to reach its normal location in
the interventricular sulcus for the left coronary artery
or the right atrioventricular sulcus for the right
coronary. The patient may suffer exertional angina
when the two great arteries dilate because of increased cardiac output. The lesions, although rare,
are important to the radiologist because of the everincreasing use of contrast-enhanced CT of the chest.
Visualization of the main coronary arteries is usually
adequate on these studies to demonstrate the origin
of the proximal portions of the coronaries. Anomalous lesions are easily detected, if kept in mind
(Fig. 18).
Summary
Because of the addition of the ever-increasing
number of adults with corrected congenital heart
disease, in addition to the relatively stable number
of patients reaching maturity without operation, it is
likely that many radiologists will be involved with
their care. Not uncommonly, no pertinent clinical
information is provided at the time, especially when
the patient is seen at an emergency room because of
intercurrent disease. The radiologist interpreting the
films, usually chest films or CT examinations, should

be conversant with the patterns of congenital heart
disease and their postoperative appearance and potential complications.
[17]
[18]
References
[1] Moodie DS. Adult congenital heart disease. Curr Opin
Cardiol 1994;9:137 42.
[2] Warnes CA, Liberthson R, et al. TaskForce I: the
changing profile of congenital heart disease in adult
life. J Am Coll Cardiol 2001;37:1170 5.
[3] Baron MG. Plain film diagnosis of common cardiac
anomalies in the adult. Radiol Clin N Am 1999;37:
401 20.
[4] Steiner RM, Reddy GP, Flicker S. Congenital cardiovascular disease in the adult: imaging update. J Thorac
Imaging 2002;17:1 17.
[5] Roberts WC. The congenitally bicuspid aortic valve: a
study of 85 autopsy patients. Am J Cardiol 1970;26:
72 83.
[6] Siltanen P. Atrial septal defect in adults: clinical and
haemodynamic studies of 128 cases before and after
surgical correction under cardiopulmonary bypass.
Acta Med Scand Suppl 1968;497:1 151.
[7] Steele P, Fuster V, Cohen M, et al. Isolated atrial septal
defect with pulmonary vascular obstructive disease:
long-term follow-up and prediction of outcome after
surgical correction. Circulation 1987;76:1037 42.
[8] Granton JJ, Rabinovitch M. Pulmonary arterial hypertension in congenital heart disease. Cardiol Clin 2002;
20:441 57.
[9] Campbell M. Natural history of coarctation of the aorta. Br Heart J 1970;32:633 40.
[10] Cohen M, Fuster V, Steele PM, Driscoll D, McGoon
DC. Coarctation of the aorta: long term follow-up and
prediction of outcome after surgical correction. Circulation 1989;80:840 5.
[11] Toro-Salazar OH, Steinberger J, Thomas W, et al.
Long-term follow-up of patients after coarctation of
the aorta repair. Am J Cardiol 2002;89:541 7.
[12] Vlodaver Z, Neufeld HN. The coronary arteries in coarctation of the aorta. Circulation 1968;37:449 54.
[13] Von Kodolitsch Y, Aydin MA, Koschyk DH, et al.
Predictors of aneurysmal formation after surgical correction of aortic coarctation. J Am Coll Cardiol 2002;
39:617 24.
[14] Bromberg BI, Beekman RH, Rocchini AP, et al. Aortic
aneurysm after patch aortoplasty repair of coarctation:
a perspective analysis of prevalence, screening tests
and risks. J Am Coll Cardiol 1989;14:734 42.
[15] Mendelsohn AM, Crowley DC, Lindauer A, Beekman
III RH. Rapid progression or aortic aneurysms after
patch aortoplasty repair of coarctation of the aorta.
[16] Niwa K, Perloff JK, Bhuta SM, et al. Structural abnormalities of great arterial walls in congenital heart
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
689
disease: light and electron microscopic analysis. Circulation 2001;103:393 400.

Netter F, van Mierop LHS. Embryology. In: Yonkman
FY, editor. Heart. Summit (NJ): Ciba Pharmaceutical;
1969. p. 119 24.
Lillehei CW, Cohen M, Read RC, et al. Direct vision intracardiac surgical correction of the tetralogy
of Fallot, pentalogy of Fallot and pulmonary atresia
defects: report of the first ten cases. Ann Surg 1955;
142:418 45.
Therrien J, Marx GR, Gatzoulis MA. Late problems in
tetralogy of Fallot: recognition, management, and prevention. Cardiol Clin 2002;20:395 494.
dUdekem Y, Ovaert C, Grandjean F, et al. Tetralogy of
Fallot: transannular and right ventricular patching
equally affect late functional status. Circulation 2000;
102(Suppl 3):III-116 III-22.
Ilbawi MN, Idriss FS, DeLeon SY, et al. Factors that
exaggerate the deleterious effects of pulmonary insufficiency on the right ventricle after tetralogy repair.
J Thorac Cardiovasc Surg 1987;93:36 44.
Jonsson H, Ivert T, Jonasson R, et al. Work capacity
and central hemodynamics thirteen to twenty-six years
after repair of tetralogy of Fallot. J Thorac Cardiovasc
Surg 1995;110:416 26.
dUdekem Y, Pasquet A, LeBreux L, et al. Does right
ventricular outflow tract damage play a role in the
genesis of late right ventricular dilatation after tetralogy of Fallot repair? Ann Thorac Surg 2003;76:555 61.
Reddington AN. Right ventricular function. Cardiol
Clin 2002;20:341 9.
Davlouros PA, Kilner PJ, Hornung TS, et al. Right ventricular function in adults with repaired tetralogy of Fallot assessed with cardiovascular magnetic resonance
imaging: detrimental role of right ventricular outflow
aneurysms or akinesia and adverse right-to-left ventricular interaction. J Am Coll Cardiol 2002;40:2044 52.
Niwa K, Sin SC, Webb GD, Gatzoulis MA. Progressive aortic root dilatation in adults late after repair of
tetralogy of Fallot. Circulation 2002;106:1374 8.
Roman MJ, Devereux RB, et al. Two-dimensional
echocardiographic aortic root dimensions in normal
children and adults. Am J Cardiol 1989;64:507 12.
Dodds III GA, Warnes LA, Danielson GK. Aortic
valve replacement after repair of pulmonary atresia
and ventricular septal defect or tetralogy of Fallot.
J Thorac Cardiovasc Surg 1997;113:736 41.
Liebman J, Cullum L, Belloc NB. Natural history
of transposition of the great arteries: anatomy and
birth and death characteristics. Circulation 1969;40:
237 62.
Hornung TS, Derrick GP, Deanfield JE, Reddington
AN. Transposition complexes in the adult: a changing
perspective. Cardiol Clin 2002;20:405 20.
Wilson NJ, Clarkson PM, Barratt-Boyes BG, et al.
Long-term outcome after the Mustard repair for simple
transposition of the great arteries: 28-year follow-up.
Jatene AD, Fontes VF, Paulista PP, et al. Anatomic
690
correction of transposition of the great vessels. J Thorac

Cardiovasc Surg 1976;72:364.
[33] Tulevski II, van der Wall EE, et al. Usefulness of magnetic resonance imaging dobutamine stress in asymptomatic and minimally symptomatic patients with
decreased cardiac reserve from congenital heart disease
(complete and corrected transposition of the great arteries and subpulmonic obstruction). Am J Cardiol 2002;
89:1077 81.
[34] Losay J, Touchot A, Serraf A, et al. Late outcome after
arterial switch operations for transposition of the great
arteries. Circulation 2001;104(Suppl 1):I12 6.

Radiology Clinics of North America Cardiac Imaging

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Radiology Clinics of North America Cardiac Imaging

Diunggah oleh

Hak Cipta:

Format Tersedia

Radiol Clin N Am 42 (2004) xi xii

It is our pleasure to present this issue of the

MR imaging, now allows use of these imaging

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) xixii

anatomy and geometry of the cardiac silhouette. A

studies are clearly described and illustrated. This area

Radiol Clin N Am 42 (2004) 487 495

How to approach cardiac diagnosis from the

Direct visualization of pathologic changes in the

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495

Approach to the film

Box 1. A cardiac diagnostic approach to

apex, and aortic knob

Box 2. Frontal chest radiograph: normal

This approach serves well in developing a disciplined

 Heart: position, size, and shape

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495

Box 3. Lateral chest radiograph: normal

The ability to diagnose is based on the expectation of

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495

tribution, even in an individual with normal cardiac function.

Fig. 3. Coronal scan through chest obtained with a spin echo

frontal projection, are bounded by radiolucent lung

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495

Box 4. Questions to ask

main pulmonary artery? Normally

double aortic arch. A barium swallow should be

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495

in interpreting routine chest studies. See the article by

Fig. 6. This radiograph illustrates the classical findings of

and has always been asymptomatic. The aortic knob

Fig. 7. Frontal chest radiograph demonstrating enlargement

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495

Box 5. Causes of pulmonary arterial

Destructive lung disease

firms the finding. This patient is at risk of rupture

arterial hypertension, and indicates pulmonic valve

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495

Cardiac valve disease

Fig. 9. Diagram of a lateral chest radiograph indicating the

enlargement of the left atrial appendage along the left

The normal position of each of the four heart

M.J. Lipton, L.M. Boxt / Radiol Clin N Am 42 (2004) 487495

aortic valve stenosis, Marfan syndrome, and right

Position of the heart

Traditional plain chest radiography provides the

Radiol Clin N Am 42 (2004) 497 514

How to plan and perform a cardiac

Cardiovascular disease is the major cause of death

This article is a revision of an article published in Magn

Cardiac MR imaging techniques: general

M. Poustchi-Amin et al / Radiol Clin N Am 42 (2004) 497514

should be instructed that they might feel flushed or

Cardiac gating and physiologic monitoring

magnetic field and the radiofrequency pulse may

Cardiac MR imaging pulse sequences

M. Poustchi-Amin et al / Radiol Clin N Am 42 (2004) 497514

include conventional spin echo, breathhold turbo or

jet and quantification of aortic regurgitation. Most

Cardiac imaging planes

M. Poustchi-Amin et al / Radiol Clin N Am 42 (2004) 497514

lesions and may complement morphologic evaluation

Coronal and sagittal planes

Double-oblique (oblique-sagittal) planes through

Heart: position, size, and shape