Int. J. Oral Maxillofac. Surg.

1993," 22:173-177 Copyright Munksgaard 1993

Printed in Denmark. All rights reserved
InternationalJournal of
Oral&
Max~ofacialSurgery
ISSN 0901-5027

Research; development
B. Hell 1, R A. Walter 2,
Three-dimensional St. Schreiber 3, H. Blase 3,
G. Bielke 4, St. Meindl 4, G. Stein 4
~Department of Maxillofacial and Plastic
ultrasonography in maxillofacial Surgery, University Hospital Rudolf Virchow
of Free University Berlin; 2Department of
Radiology, CarRas-Klinik Rastpfuhl,

surgery Saarbr0cken; 3Medical Engineer Office,

MediCAD, Darmstadt; 4Deutsche Klinik f~r
Diagnostik, Wiesbaden, Germany

A new diagnostic tool

B. Hell, E A. Walter, St. Schreiber, H. Blase, G. Bielke, St. Meindl, G. Stein."

Three-dimensional ultrasonography in maxillofacial surgery. A new diagnostic tool
Int. J. Oral Maxillofac. Surg. 1993; 22." 173 177. Munksgaard, 1993

Abstract. Three-dimensional (3-D) ultrasonography is introduced as a new diag-

nostic method for soft-tissue lesions. This method is based on the production of
absolutely parallel brightness (B)-scan tomographies with constant intervals.
After adequate preparation, the sonographic primaries are processed into 3-D
graphs by a specialized program called EUCLID-IS. This method promises to
be useful in the evaluation of tumor patients as well as in the follow-up examination
of these patients. The main advantages of the presented method are the capacity
to measure accurately the volume of regions of interest and the excellent visualiza-
tion of the structures examined.

Zusammenfassung: Die 3-D-Sonographie wird als eine neue Untersuchungstech-

nik bei der Beurteilung von Weichgewebsl/isionen vorgestellt. Die Basis der
Technik stellen absolut achsenparallele,/iquidistante konventionelle B-Scan-
Sonographien dar. Nach einer entsprechenden Aufbereitung der prim/iren Sono-
gramme werden diese in ein dreidimensionales Computer-Aided Design (CAD)-
Programm (EUCLID-IS) transferiert. Die Methode erscheint besonders hilfreich
bei der Beurteilung von tumor6sen L/isionen inklusive der region/iren Lymphab-
Key words: B-scan sonography; 3-D imaging;
flugwege sowohl p/iroperativ wie auch in der Tumornachsorge. Die bedeutendsten volumetry; lymph nodes.
Vorteile liegen in der M6glichkeit zur Volumetrie der interessierenden Strukturen
und in einer excellenten Visualisation der untersuchten Strukturen. Accepted for publication 2 February 1993

Ultrasound evaluation of the head and
neck, especially B (brightness) mode
sonography, has been an accepted
method of examining soft tissue for
some years. The clinical importance
of this method is widely accept-
ed 1'~6'7'9-14'16'19'25'26.Its main advantages
are that it is noninvasive, dynamic,
painless, and economical. It offers high
resolution if performed with proper
equipment, i.e. the "real-time tech-
nique", and at least 5 megacycles/s ul-
trasound frequency. While most studies
have dealt with the examination of soft
tissue only, recent publications have
also presented indications for ultra-
sound examination of bony struc-
tures 1'8'18. In addition, MENDE et al. ~7
have emphasized the usefulness of 2-
dimensional (D) ultrasound evaluation
in controlling the effect of nonsurgical
tumor therapy by measuring tumor vol-
ume. However, the judgment of volumes
by the use of conventional B-scan sono-
graphy was difficult since it seemed to
be impossible to reestablish the same
"patient-scanner" relationship in a se-
cond evaluation.
Ultrasound examination also pre-
sents other difficulties. A detailed
knowledge of the sonostructure of the
region examined is necessary to recog-
nize pathologic structures. In 1990, NO-
RE R 19 published a monograph dealing
with specific sonographic findings of
anatomic sections in the head and neck
region. He proposed that standardized
views be used to guarantee complete
examination of this area. To produce a
valuable sonographic examination, the
echographist must be able to put several
sections together into a 3-D context. In
174 H e l l et al.

several cases, however, the examiner has orientation easy. The volume of interesting third section of the whole primary sono-
difficulty in d e m o n s t r a t i n g these find- anatomic or pathologic conditions can also graphic set into the computer in the above
ings to other surgeons, w h o have to rely be calculated by the program used. The infor- mentioned manner. With EUCLID-IS, the
mation can be transferred to the computer- following 3-D sonographies were produced.
on the f o r m e r ' s findings. This is due to
aided-manufacturing (CAM) program, al- An anterior-medial view of the left side
the difficulty o f interpreting s o m e iso-
lowing the production of models based on presented all marked structures of the sket-
lated s o n o g r a p h i c t o m o g r a m s . There is the sonographic scans. ches. Between the carotid artery, marked as
clearly a need for a m e t h o d t h a t presents a violet band, and the brown sternocleido-
results clearly to clinicians. mastoid muscle, the internal jugular vein to-
Case report
It is the aim o f the present study to gether with another venous branch appeared
describe a s o n o g r a p h i c m e t h o d which A 31-year-old woman presented with a large as blue structures. Between the muscle and
allows easy interpretation o f sono- mass within the left parotid gland and two the yellow-appearing skin, the green-marked
graphic findings and permits volume subcutaneous tumors on the left sternocleido- lymph nodes were situated (Fig. 3).
mastoid muscle. B-scan sonography showed To allow a realistic impression and to judge
m e a s u r e m e n t s o f the lesions diagnosed.
that the tumors displaced some soft tissues the distance between large vessels and lymph
(Fig. l a and b). Fig, 2a and b show one nodes, we "removed" the sternocleidomas-
Material and methods example of the primary set of sonographic toid muscle (Fig. 4).
sections made on the lateral cervical side. A An anterior-lateral view revealed the
The basis of 3-D sonography is the produc- total of 128 scans parallel to the common lymph node structures on the sternocleido-
tion of absolutely parallel sonographic sec- carotid artery of an area 46 mm below the mastoid muscle after computerized "re-
tions of tissues in equal intervals. In 1983, skin surface was made. These data conse- moval" of the overlying skin (Fig. 5).
BmLt(E3 described an apparatus which met quently led to a primary section thickness An anterior-medial view revealed the skin
that condition, and this machine was used in of 0.359 mm. In order to produce a tissue and lymph nodes only after "removal" of
this study. With knowledge of the number of thickness of about 1 mm, we input only each all other structures. The relationship between
performed sections and the measurement of
the body surface evaluated, the thickness of
tissue represented by one slice can be calcu-
lated. All primary ultrasound scans were pro-
cessed by one computer (Contron Electronic,
Munich, Germany).
The next step to 3-D ultrasound is the so-
called segmentation; that is, the most import-
ant structures of each scan have to be
marked. Since a conventional 2-D ultrasono-
graphic scan is built up of many pixels with
different brightnesses (B-scan sonography)
and interpositioned black spaces, these tomo-
grams do not allow continuous information.
Therefore, the production of computer-aided
segmentation is not available at present. Each
segmentation was therefore made by hand.
Precautions were taken to ensure that the
enlargement of the y- and x-axis of each seg-
mented sonographic primary given by its two
dimensions was also transferred to the z-axis,
Fig. 1. B-scan sonography of superior lymph node in neck: a) section parallel to carotid
which was built up by the total number of
artery: 4 14 mm in size lymph node presents as hypoechoic lesion lying on sternocleidomastoid
all sections. In addition, the orientation of
muscle; b) section parallel to clavicle: second plane shows hypoechoic lymph node 5-12 mm
the segmentations had to be well defined.
in size between skin and sternocleidomastoid muscle.
The sketches produced by segmentation
were then put into a second computer as the
third step to 3-D sonography. A computer
program called EUCLID-IS (MATRA Data-
vision, Darmstadt, Germany) was used for
3-D reconstruction of the sonographic 2-D
scans. The program used was originally intro-
duced to solve general industrial construction
problems; for example, those of airplane or
car production. The program used could deal
with scans representing about 1 mm of tissue
thickness. According to the previously deter-

.aot I orno
mined tissue thickness of one slice, all 2-D
scans of one primary sonographic set were
put together in a 3-D space, a procedure
which enables exact calculation of volume
measurements of lesions. By this method, all 2b
segmented bodies could be shown three-di-
mensionally; for example, vessels, muscles, Fig. 2. Example of set of 128 B-scan sonographies parallel to carotid artery. Scan shows
glands, or lymph nodes. The program also largest dimension of removed lymph node. Volume of examined region lies under skin surface
allows selection of different colors for each of 46 mm 2. Distance between each section is constant: 0.359 mm. a) original sonographic
anatomic or pathologic structure, making scan; b) sketch of same section.
 Three-D ultrasonography 175

Fig. 3. 3-D sonography of left neck including pathologic lymph node structures (anterior ventral view). Each edge of blue cube is 1 cm.
Carotid artery: violet band; internal jugular vein and venous branch: blue; sternocleidomastoid muscle: brown; lymph nodes: green.
Fig. 4. Subtraction of muscle, revealing distance between lymph nodes and vessels.
Fig. 5. Computerized "operating" on screen (simulated removal of the skin): lymph nodes are situated directly on muscle.
Fig. 6. Relation of lymph nodes to skin remaining after all other structures have been removed. This view can be decisive in planning surgical
procedures.
Fig. 7. View from top into neck, revealing close connection between lymph nodes, skin, and muscle.
Fig. 8. 3-D view of single lymph node by virtual light.

lymph nodes and skin could be judged (Fig. lymph nodes to both skin and muscle was the anatomic situation was possible. The
6). demonstrated (Fig. 7). computerized measurement of the caudal
In a view directed from the head to the With the use of a virtual light source and lymph node revealed a volume of 499 mm 3
lateral shoulder, the close relationship of the the resulting reflections, a realistic view of (Fig. 8). This lymph node was surgically re-
176 H e l l et al.
Fig. 9. Outer surface of model of ascending
ramus of mandible produced by CAD-CAM
system using 3-D sonography and laserlitho-
graphy.
moved. The volume was 500 mm3. Histologic
examination confirmed the diagnosis of a
highly malignant lymphoma.
Discussion
Three-dimensional sonography was
proposed as early as 195C 5. HowRY et
al. ~5favored a method based on stereo-
scopic sonographic views. In 1965,
BAUM2 presented another modification
of 3-D sonography: parallel sections of
primary sonographic scans were put to-
gether by an artist, who made drawings
as he visualized the 3-D figure of the
region of interest from these primary
scans. These results were not further de-
veloped. In the late 1980s, SOHN et
al. 2-24 published some studies dealing
with 3-D sonography. He used a pencil-
like section scanner which emitted ultra-
sound waves into the tissue in a diver-
gent manner. Therefore, only a small
area of tissue was superficially evalu-
ated, while at depth the examined vol-
ume was enlarged. This method allowed
a useful 3-D ultrasound examination
only in deeper regions of the body. Since
SOHN was a gynecologist interested in
examining early pregnancies, he used
this method successfully. However, in
maxillofacial surgery, where more
superficially located structures have to
be evaluated, this method is not useful.
The method of SOHN et al. included also
a rotation of the ultrasound probe at
the same place. Every 10 a new scan
was produced. Thus, he needed only 18
pictures for his 3-D reconstruction; as
a consequence, only a rough 3-D picture
was produced.
Our method allows a 3-D reconstruc-
tion of even subcutaneous structures
and at the same time offers the capacity
to measure volumes in a reproducible
manner. The method is a modification
of BAUM'S principle2 by modern com-
puter programs.
Because of the above mentioned
capabilities, the method is indicated for
the following purposes:
l) to determine whether nonsurgical
tumor therapy is effective; that is,
during chemotherapy and/or ir-
radiation, and at follow-up examin-
ations after tumor therapy
2) to measure alterations in volume of
structures under examination in
cases indicated.
The excellent visualization of sono-
graphic findings can be useful in plan-
ning surgical procedures. For example,
3-D sonography could help to deter-
mine the quantity of skin or bone to be
resected and to select the reconstructive
technique. The method described also
allows different views of regions of
interest which are not possible by other
methods (Fig. 7). This might also be
advantageous in assessing lesions.
The method described allows reevalu-
ation of 2-D-scan sonography without
the patient being present, because paral-
lel sections in known distances are pro-
duced. The primary sonographic data
of our examinations are collected ac-
cording to NORER~9, who described the
sonoanatomy of the head and neck re-
gion in detail and determined scan di-
rections. Thus, a well-defined sono-
graphy is made accessible to more ex-
perts in this field.
Costs for storage of the primary scans
are remarkably low, because the usual
video tape is sufficient, and many sono-
graphic sets can be stored on one tape.
Data of the computer-assisted design
(CAD) system can also be transferred
to the CAM system. CAM offers the
capability to produce models of the
CAD data, leading to a model of the
examined region. Fig. 9 shows the outer
surface of the ascending ramus of the
mandible produced by the CAD-CAM
system with laserlithography. Since
bony structures cannot be penetrated
by sonographic waves, only the bony
surface directed to the scanner can be
evaluated. Because the examination of
soft tissues presents information on the
total tissue surface examined, a "volume
model" of the region of interest is poss-
ible, making surgical planning some-
what easier.
The transformation of sonographic
data into the EUCLID-IS system was
performed by means of drawings made
by hand from the sonographic screen.
A method is now being developed which
allows a direct data transfer from the
sonographic screen into the EUCLID-
IS system. Another area of development
is an automatic or at least semiautoma-
tic segmentation of the reproduced sec-
tions.
In conclusion, 3-D sonography has
proven to be a valuable adjunct to the
examination of the OMF surgeon.
However, the method requires some im-
provements to make it useful in daily
practice. Further developments are
underway which will probably lead to
wide application of this system.
References
1. AKIZUKIH, YOSHIDAH, MICH~K. Ultra-
sonographic evaluation during reduction
of zygomatic arch fractures. J Cranio-
Max-Fac Surg 1990: 18: 263-6.
2. BAUMG. A synopsis of ophthalmic ultra-
sonography. Wiss Z Humboldt-Univ
Math-Nat R 1965: 14:51 62.
3. BIELKE G. Entwicklung eines Verfahrens
zur sonographischen "in vivo" Differen-
zierung menschlichen Gewebes. Disser-
tation Fachbereich 18 (Physik) der Johan-
nes Gutenberg-Universitfit in Mainz,
1983.
4. BRUNETONJN. Ultrasonography of the
neck. Berlin: Springer, 1987.
5. BRUNETON JN, CARAMELLA E, BOUBLIL
JL, Roux P, ABBESM, DEMARDF. Echo-
graphic aspects of thyroid and para-
thyroid localizations in Non-Hodgkin
lymphomas. Fortschr R6ntgenstr 1982:
136: 530-3,
6. BRUNETONJN, ROUX P, CARAMELLAE,
DEMARD F, VALLICIONI J, CHAUVEL P.
Ear, nose, and throat cancer: ultrasound
diagnosis of metastasis to cervical lymph
nodes. Radiology 1984: 152: 771-3.
7. CHADOSHPL, SILBREYR, OEN KT. Diag-
nostic use of u~rasound in diseases of the
head and neck. Laryngoscope 1980: 90:
814~21.
8. FRIEDRICH RE, VOLKENSTEINRJ. Wertig-
keit der Sonographie ffir die Diagnose
yon Jochbogenfrakturen. Dtsch Z Mund
Kiefer Gesichtschir 1991: 15: 4724.
9. HAUENSTEIN H, ROTHE F, STEINKAMP B.
Ultrasonographische Untersuchungen
bei Tumorexl im Mundboden-Zungen-
Bereich und bei computertomographisch
gesicherten Halslymphknotenmetastasen.
Dtsch Zahnfirztl Z 1981: 36: 746-51.
10. HELL B. B-Scan-Sonographie in der
Mund- Kiefer- und Gesichtschirurgie.
Fortschr Kiefer Gesichtschir 1987: 32:
133 5.

11. HELL B. Bedeutung der Ultraschalldiag-
nostik bei Speicheldrfisenerkrankungen.
Dtsch Z Mund Kiefer Gesiehtschir 1988:
12: 318-23.
12. HELL B. B-scan-sonography in maxillo-
facial surgery. J Cranio-Max-Fac Surg
1989: 17: 39-45.
13. HELL B. Atlas der Ultraschalldiagnostik
im Kopf-Hals-Bereieh. Stuttgart: Thieme,
1990.
14. HELL B, WALTER FA. Problematik der
Beurteilung sonographischer Untersuch-
ungen. Dtsch Z Mund Kiefer Gesichts-
chir 1988: 12: 423-6.
15. HDWRY DH, POSAKONY G, CUSHMANR,
HOLMES JH. Three-dimensional and
stereoscopic observation of body struc-
tures by ultrasound. J Appl Physiol 1956:
9: 304-6.
16. MANN W. Ultraschall im Kopf-Hals-Be-
reich. Berlin: Springer, 1984.
17. MENDE U, Z6LLER J, DROMMER RB,
REICHARDT P, BORN IA. Optimization of
preoperative chemotherapy by ultra-
sound. Int J Oral Maxillofac Surg 1990:
19: 363-6.
18. MILLESI W, PRAYER L, HELMER M,
GRITZMANNN, Diagnostic imaging of tu-
mor invasion of the mandible. I n t 3 Oral
Maxillofac Surg 1990: 19: 294-8.
19. NORER B. B-Scan-Sonographie des
Mundbodens, der Wange und des oberen
Halses. Stuttgart: Thieme, 1990.
20. SOHY C, GROTEPASS J, SCHNEIDER W, et
al. Dreidimensionale Darstellung in der
Ultraschalldiagnostik. Erste Ergebnisse.
Dtsch Med Wochenschr 1988a: 113:
1743-7.
21. SOHY C, GROTEPASSJ, SCHNEIDER W, et
al. Erste Untersuchungen zur dreidimen-
sionalen Darstellung mittels Ultraschall.
Z Geburtshilfe Perinatol 1988b: 192:
241-8.
22. SOHN C, GROTEPASSJ, MENGE KH, AME-
LINGW. Klinische Anwendung der dreidi-
mensionalen Ultraschalldarstellung.
Dtsch Med Wochenschr 1989a: 114:
534-7.
Three-D ultrasonography 177
23. SOHN C, RUDOFSKY G. Die dreidimen-
sionale Ultraschalldiagnostik ein neues
Verfahren ffir die klinische Routine? U1-
traschall Klin Prax 1989b: 4: 219-24.
24. SOr~NC, STOLZW, NUBERB, et al. Verbes-
serungen der 3-D-Ultraschalldarstellung.
Bildgebung 1991: 58:116-20.
25. WALTER FA, HELL B. Die Bedeutung der
B-Scan-Sonographie in der Kopf- und
Halschirurgie. Dtsch Z Mund Kiefer Ge-
sichtschir 1985: 9:207 12.
26. ZAHN W~ KUHLO W, MACHTENS E. Ein-
satzm6glichkeiten der Ultraschalldiag-
nostik auf dem Gebiet der Mund- Kiefer-
Gesichtschirurgie. Dtsch Z Mund Kiefer
Gesichtschir 1983: 7: 177-82.
Address:
Berthold Hell, MD, DDS
Universitiitsklinikum Rudolf Virchow
Augustenburger Platz, Wedding
1000 Berlin
Germany