Anda di halaman 1dari 14

Bioprinting 3 (2016) 1–14

Contents lists available at ScienceDirect

Bioprinting
journal homepage: www.elsevier.com/locate/bprint

A review on design for bioprinting


a,b,c,d,⁎ a
crossmark
Ibrahim Ozbolat , Hemanth Gudapati
a
Engineering Science and Mechanics Department, Penn State University, University Park, PA 16802, USA
b
Biomedical Engineering Department, Penn State University, University Park, PA 16802, USA
c
The Huck Institutes of the Life Sciences, Penn State University, University Park, PA 16802, USA
d
Materials Research Institute, Penn State University, University Park, PA 16802, USA

A R T I C L E I N F O A BS T RAC T

Keywords: In order to bioprint living tissue and organ constructs, patient-specific anatomical models need to be acquired;
Design requirements for bioprinting however, these models mainly provide external surface information only. The internal architecture of tissue
Medical imaging constructs plays a crucial role as it provides a porous environment for media exchange, vascularization, tissue
Blueprint modeling growth and engraftment. This review presents design requirements for bioprinting and discusses currently
Toolpath planning
available medical imaging techniques used in acquisition of anatomical models including magnetic resonance
imaging (MRI) and computed tomography (CT), and compares their strengths and limitations. Then,
consideration for design architecture is discussed and various approaches in blueprint modeling of tissue
constructs are presented for creation of porous architectures. Next, existing toolpath planning approaches for
bioprinting of tissues and organs are presented. Design limitations for bioprinting are discussed and future
perspectives are provided to the reader.

1. Introduction Despite the wide array of biofabrication techniques used in


manufacturing of tissue constructs, most do not allow generation of
Tissue engineering, an interdisciplinary field of biology, biomater- highly intricate geometries including the anatomical shape of the tissue
ials and engineering, is seeking to restore tissue functions by develop- as well as the complex porous internal architecture [5]. Thus, 3D
ing engineered three-dimensional (3D) tissue constructs providing the bioprinting offers several advantages over traditional scaffold fabrica-
optimum environment for cell attachment and growth, tissue regen- tion techniques (i.e., fiber bonding, solvent casting/particulate leach-
eration, fluid movement and structural integrity [1]. Engineered tissue ing, gas foaming and molding [6,7]) [8]. One of the major advantages
constructs attempt to mimic both the anatomical shape (external of 3D bioprinting is that patient-specific information can be directly
geometry) and opposite impression of internal architecture of replaced incorporated into the biofabrication process in order to generate
tissues. The acquisition of the external geometry of tissue constructs is anatomically-correct shapes. Bioprinting has the ability to co-culture
critical in order to precisely represent anatomically-correct shapes [2]. multiple cell types locally and facilitates their precise patterning.
In regenerative medicine, non-invasive medical imaging techniques Besides, bioprinting enables controlled delivery of growth factors and
including, but not limited to, computed tomography (CT) and magnetic genes, and allows fabrication of tissue models in a high-throughput
resonance imaging (MRI) techniques are widely used for acquisition of manner [9]. Using bioprinting, porous tissue constructs can be built in
patient-specific information [3]. Upon acquisition of the 3D model of high resolution with the proper design to provide the appropriate
tissues and organs from the patient, an internal architecture needs to oxygenation for cellular components for successful regeneration of
be designed, which includes internal channels and interconnected tissues and organs [10].
pores enabling cell attachment, cell proliferation, nutrient flow and In addition to facilitating cellular activity post-bioprinting, design
tissue regeneration [4]. The internal architecture and topology has a of tissue constructs must incorporate a tolerance of mechanical load,
substantial influence on growth and proliferation of cells; hence, an mechanical and biochemical stimulation as well as flexibility in shape
optimum design is needed to accelerate tissue regeneration. If the and size to accommodate growth within the patient (i.e., tissue
designed tissue constructs are developed for transplantation purposes, constructs for pediatric cases). For instance, the majority of bone
they must possess the anatomically correct shapes of the tissues and tissue undergoes mechanical loading so a bone construct must be
organs from individual patients and they should engraft with the host. designed to withstand mechanical stresses while the tissue is generat-


Corresponding author at: Engineering Science and Mechanics Department, Penn State University, University Park, PA 16802, USA.
E-mail address: ito1@psu.edu (I. Ozbolat).

http://dx.doi.org/10.1016/j.bprint.2016.11.001
Received 13 July 2016; Received in revised form 29 October 2016; Accepted 8 November 2016
Available online 11 November 2016
2405-8866/ © 2016 Elsevier B.V. All rights reserved.
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

Fig. 1. The major steps involved in bioprinting design including medical imaging, blueprint modeling, toolpath planning and 3D bioprinting.

ing within the constructs. Similarly, during regeneration of articular manufacturing limitations [15]; however, the porosity level should
cartilage tissue, dynamic loads are applied to the implanted constructs vary spatially based on mechanical, biological and functional require-
during tissue repair. Thus, the designed tissue construct must with- ments. For example, fabrication of vertebral bone constructs with
stand the dynamic loads until complete regeneration takes place. uniform material and structural architecture is not a reasonable
Tissue constructs must also satisfy biological requirement such as cell approach. Natural bones exhibit functional gradients throughout their
attachment and proliferation, cell signaling and transport of nutrients structures. The load exerted on a native vertebra is distributed more to
and metabolic waste products [11]. Thus, mechanical, biological and the outer regions compared to inner sites. Moreover, cell loading and
biochemical characteristics of the regenerating tissue environment proliferation starts from the exterior side of the bone construct and cell
should inform the design of tissue constructs prior to bioprinting. ingrowth proceeds through the core. This necessitates higher porosity
Fig. 1 shows the major steps involved in design of tissue constructs in the exterior regions and lower porosity in the core. High porosity
for 3D bioprinting. First, medical images obtained from MRI and CT supports more cell growth and proliferation but limits integrity and
scans coupled with alternative approaches (i.e., reverse engineering) load bearing capacity of tissue constructs, thus porosity levels in
can be used to generate a 3D model of the tissue and organ to be distinct regions of the construct should be appropriate for the func-
replaced. Next, various blueprint modeling approaches, such as tional requirements of each region. In this regard, Hollister et al. [16–
computer-aided design (CAD)-based systems, image-based systems, 18] used an image-based homogenization method integrated within a
freeform systems, implicit surfaces and space filling curves, are topology optimization approach to design microstructural and material
instrumental in the design of the internal architecture of tissue properties for tissue constructs, matching elastic properties and
constructs. Then, toolpath algorithms are used to generate path plans porosity of bone tissue simultaneously. Porosity and effective stiffness
in Cartesian or parametric form. Finally, different 3D bioprinting properties were optimized within theoretical boundaries with the
modalities can be used to fabricate tissue constructs. objective of mimicking the properties of natural tissues. Required
porosity was ensured for a wide range of permeability to enable cell/
gene delivery with suitable stiffness properties to human bone tissue by
2. Design requirements for 3D bioprinting
choosing the true biomaterial and unit cell architecture together.
Wettergreen et al. [19,20] addressed the porosity determination in
In regenerative medicine, synthetic engineered implantable tissues
each layer by mapping them with CT scan images and achieving the
and organs should have functional gradients to mimic their original
desired porosity using unit cells from a library as building blocks. In all
counterpart in addition to conformational geometries that have been
of the abovementioned studies, 3D printing was used to fabricate tissue
extensively studied in the literature [12]. Engineered tissue constructs
constructs and achieved improved mechanical properties.
attempt to mimic both the external geometry and opposite impression
In addition to creating pores, there are also efforts to create actual
of internal architecture of replaced tissues. Cells loaded in tissue
channels that attempt to mimics the microvasculature. Two approaches
constructs require nutrients, proteins, growth factors and waste
have been used in the literature for patterning channels for mimicking
removal, making mass and fluid transport vital to cell survival. The
microvasculature networks. The first approach involves direct bioprint-
size, geometry, orientation, interconnectivity and surface chemistry of
ing of the vasculature network using a coaxial nozzle apparatus [21–
pores and channels determine the nature of nutrient flow [13].
23]. The second approach is indirect bioprinting, which is bioprinting
Moreover, interconnectivity of pores has a direct impact on cell
of a sacrificial polymer that is removed by thermally induced decros-
migration and tissue ingrowth. Thus, a well-connected internal net-
slinking, leaving a vascular network behind [24]. For more detailed
work of channels is vital to the survival and development of the
discussion, the reader is referred to our earlier review [25].
generated tissue, allowing deeper fluid and nutrient transport and
Table 1 presents the design factors affecting tissue construct
release of biomaterials and biomolecules. An optimal engineered tissue
properties such as mechanical, biological, and geometric, transport
design incorporates these features to accelerate cell growth and
and fabrication. All these properties affect the performance of the
proliferation.
construct upon implantation. During scaffold design and fabrication,
Determining the optimum porosity of designed constructs is crucial
biomaterial selection, internal architecture, porosity level and inter-
as the porosity mediates cell attachment and growth both in vitro and
connectivity, external geometry, permeability and compatible bioprint-
in vivo. Although high porosity supports cell growth and tissue
ing processes must be carefully considered for enhanced tissue healing
generation, it limits the integrity and load bearing capacity of tissue
to occur. In addition to design factors, fabrication of tissue constructs is
constructs; thus, the optimum porosity lies within a critical range [14].
also of considerable importance in tissue engineering, where operating
The porosity distribution is a multi-objective criterion to meet biolo-
and environmental conditions, control mechanisms, resolution and
gical and mechanical requirements simultaneously. Therefore, internal
bioink selection are key factors in the process. Three major bioprinting
architecture should be designed with different regions having appro-
modalities, including extrusion-based bioprinting (EBB) [25], droplet-
priate porosity levels in order to meet abovementioned requirements.
based bioprinting (DBB) [26] and laser-based bioprinting (LBB) [27],
The majority of tissue constructs have been designed with constant
have been used to fabricate 3D tissue constructs using a layer-by-layer
porosity level throughout their structure because of design and

2
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

Table 1 tion), which enables complex patterning [38,40]. However, collagen


Tissue construct design consideration and parameter selections affecting tissue construct type I crosslinks slowly [41] and the support through buoyancy is
properties (modified from [118]).
difficult because of the spreading issues requiring alternative strategies
Properties Design consideration Selections affecting for complex patterning. Thus, bioink characteristics affect the scaffold
tissue construct design and fabrication strategies substantially. Overall, software pro-
properties grams or algorithms should strategize toolpath and bioprinting process
parameters based on rheological properties of bioink materials.
Mechanical Structural integrity, internal Bioink selection, internal
architecture stability, strength and architecture, porosity and
stiffness pore distribution, 3. Medical imaging
bioprinting modality
Biological Cell loading, cell distribution, Layout, pore size, In order to design tissue constructs (i.e., tissue scaffolds, and tissue
nutrition, cell attachment and interconnectivity,
and organ substitutes), geometric properties of the constructs should
growth, cell viability and vasculature, cell density,
differentiation, cell-cell and cell- bioink selection, bioprinted be acquired to generate accurate and anatomically-correct representa-
matrix interactions, cell cell types tion of the native tissues and organs. There are currently a number of
aggregation and tissue formation non-invasive medical imaging techniques used for imaging live tissues
Geometric Anatomical fit, tissue topology External geometry, tissue
and organs. The most common ones include MRI and CT, and are used
density
Transport Nutrient and oxygen delivery, Interconnectivity and to represent a patient's anatomy for tissue engineering and regenera-
waste removal, growth factor and permeability, bioink tive medicine as well as micro-CT (μCT), micro-MRI (μMRI), ultra-
drug delivery selection sound and nuclear medicine in order to image regenerating tissues and
Bioprinting Environmental conditions during Bioink and bioprinting organs. Although these imaging technologies can be used to obtain
bioprinting, bioprinting modality selection,
patient-specific information, sample models are currently available for
parameters, control and resolution bioprinter
most of tissues and organs on an online repository established by the
National Institutes of Health (NIH). The NIH 3D Print Exchange
deposition scheme. Extrusion-based bioprinting uses mechanical- or platform provides researchers with scientifically accurate, high-quality
pneumatic-driven system to deposit cells in the form of a filament, 3D models in a ready-to-print format [42].
whereas DBB utilizes, thermal-, piezo-, electrostatic-, electrohydrody-
namic jetting-, micro-valve- (solenoid) or acoustic-driven mechanisms 3.1. Magnetic resonance imaging
to deposit droplets of cell suspension in a high-throughput manner
[28]. Laser-based bioprinting on the other hand utilizes laser energy to Magnetic resonance imaging is a sensitive method to monitor
pattern cell-laden photopolymerizable monomers by polymerizing structural and functional changes in tissues such as tissue regeneration
them [29,30] or alternatively to deposit cells from a donor slide to a and death. Such changes are reflected in images via local variations in
receiver slide without using a nozzle [31]. Extrusion-based bioprinting tissue hydration, its physical state (e.g., freely diffusing or protein
has been the most common bioprinting modality due to ease of bound) and variations in nuclear magnetic resonance times [43]. It
operability, scalability and availability of a wide range of compatible utilizes pulsed radio frequency electromagnetic waves to excite and
bioink materials [25]. In fabrication of porous tissue constructs using generate a detectable radio-frequency signal from hydrogen atoms,
EBB, material deposition starts and stops periodically depending on which are abundantly present in human body, particularly in water and
the internal architecture, which results in a non-uniform deposited fat [44]. Changes in the parameters of pulse sequence in radio waves
layer. Therefore, the total number of starts and stops during the alter nuclear spin energy transition of the protons and hence generate
deposition process needs to be minimized. Traditional unit cell magnetic field gradients, which in turn leads to contrast among the
applications in computer-aided tissue engineering [12] limit the generated images. The generated radio-frequency signals are picked up
connectivity of internal network channels resulting in congestions in by a magnetic resonance coil and transferred to computer software for
nutrient flow. Due to the shortcomings of current computer-aided image generation.
design (CAD) technologies in representations of such micro-structures Magnetic resonance imaging has been widely used in scanning of
or the inability to transfer microstructure information to bioprinters, soft tissue components in human body [45]. Since no ionizing radiation
enhanced pore architecture with interconnected channels needs to be is involved, MRI is a preferred imaging method. The highest resolution
addressed in the bioprinting process to provide better functionality and achieved in 3 T (T) is 250 µmx250 µmx0.5 mm with a scan time of 5–
nutrient flow [13]. 40 min [46]. A higher spatial resolution (5–200 µm) can be achieved
Scaffolds for different tissue types require specific bioink materials, with μMRI in very high strength of magnetic fields such as 7–9 T [3],
which in turn influences scaffold design and fabrication strategies. where the contrast between soft tissues can be further improved by
Hard tissues, such as bone, may require bioink materials containing using contrast agents (i.e., magnetic nanoparticles [47]). Anatomical,
bioactive ceramics such as hydroxyapatite (HA) and binders such as functional and cellular information for various soft tissues can be
poly(vinyl alcohol) (PLA) [32]. Contrastingly, soft tissues, such as skin, acquired using MRI due to its superior ability to distinguish soft tissue
require bioink materials made of hydrogels (i.e., collagen, hyaluronic contrast.
acid, etc.) [33,34]. Furthermore, viscous bioink materials are pattern- Humans cannot withstand higher levels of magnetic fields, which
able with EBB and LBB, [25,35] whereas less viscous bioink materials can generate discomfort and sensations of vertigo [48]. In general, the
are patternable with DBB [26]. Different bioink materials may also magnetic field that humans are exposed to does not exceed 3 T. Longer
require different fabrication strategies even when using the same scan times can improve the resolution of images; however, some
bioprinting modality as rheological properties and the crosslinking patients can still experience anxiety and claustrophobia under longer
mechanism of the bioink determine its mechanical strength, how it is scans.
dispensed and how it interacts with the substrate [25,26,36–38].
Spreading of bioprinted droplets or filaments on the substrate causes 3.2. Computed tomography
them to deviate from their actual desired positions resulting in
structural failure of the scaffold. For example, sodium alginate cross- Computed tomography, also known as computer-axial tomography,
links rapidly with calcium chloride [39] and the bioprinted scaffold can is a non-invasive medical scanning technique that has been widely used
be supported through buoyancy (provided by calcium chloride solu- to image hard tissue components in the human body [49,50]. It utilizes
the principle of two-dimensional (2D) X-ray imaging [51], where an X-

3
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

ray tube emitting a conic beam of electromagnetic radiation selectively commercially available programs such as Materialise Mimics, Amira
penetrates the body; hard tissue (i.e., bone) attenuates X-rays much and Avizo 3D [62]. Upon loading images onto the software, the region
more than soft tissue. A detector in the opposite side of the scanner is of interest is selected and the approximate threshold value is deter-
used to acquire the cross-sectional scans of the tissue and the resulting mined in order to best capture the tissue anatomy and bioinformatics.
data is then processed to generate a cross-section of the region of Next, all pixel values within the region of interest are merged to a color
interest in human body. In order to generate 3D images, a rotational mask using region growing techniques. Although this approach is the
apparatus is used to turn around the scanned body with an angle of most common segmentation practice used in medical image proces-
increments from 0.25 to 1° and covers a scan of 360° producing 360– sing, it does not capture accurate information over a large volume of
1440 images. The scan images are then stacked using tomographic tissue samples as tissue samples may exhibit heterogeneity. Therefore,
reconstruction algorithms [52]. a homogenization approach, as demonstrated in [12], can be used to
Computed-tomography can generate higher resolution images such better capture the tissue properties over a large volume. In this
as 0.24–0.3 mm in relatively short scan times but the utilized ionized approach, the region of interest can be divided into a number of
radiation presents a finite risk to patients [46]. Patients can only be subregions, where different threshold values are assigned depending on
subjected to CT-scans at a limited dose at controlled frequencies. the tissue properties. The segmentation processes can then be per-
Computed-tomography imaging is not sensitive to metal components formed independently followed by combination of all subregions for
so that patients with implants and medical devices with batteries may representation as a single 3D object. Therefore, the homogenization
be safely subjected to the CT-scanning process. Moreover, fewer user approach provides an accurate representation of the native tissues.
manipulations are required in the acquisition of 3D medical images For a detailed discussion on image segmentation, the reader is
compared to MRI. Computed tomography has been utilized in imaging referred to earlier review articles [63,64].
bone and tumor tissue, and can differentially display soft tissue along
with the boundaries of bone tissue. 4. Blueprint modeling
In addition to CT, μCT has been used in tissue engineering,
particularly to monitor tissue regeneration in small animals such as The generated 3D model as a result of image segmentation is
rodents [53]. Micro-CT enables high resolution from 1 to 200 µm for a usually represented in stereolithography (STL) format along with other
limited volume of samples and is not considered invasive for animals formats such as but not limited to virtual reality modeling language
larger than mice [46]. Contrast agents (i.e., iodine- and nanoparticle- (VRML) and 3D graphics (X3D) [42]. Here, the STL file format should
based agents) can be administrated to improve the quality of the image not be confused with the stereolithography process, a widely used 3D
and register the soft tissue components [54]. printing technique [65]. While stereolithography technique was the
first 3D printing process invented in 1980s [66], the STL file format
3.3. Other imaging techniques emerged as a native file format to the Stereolithography CAD software.
STL is an exchange file format widely used in various applications, such
Although MRI and CT can be used in acquiring image data for as 3D printing, 3D scanning and finite element analysis [67]. Currently,
scaffold design, they can also be used in assessing the integration of a the majority of 3D printers and recent bioprinters accept input files in
bioprinted scaffold once implanted. Ultrasound imaging can also be the STL format. Upon generation of patient-specific tissue or organ
used for such a purpose [55,56]. One of the major advantages of models in a digital form, the 3D surface model is further processed to
ultrasound is the scan time, which allows for the capture of real-time generate an internal architecture to project the anatomy of regenerated
images of moving organs and blood flow in vessels. In this regard, tissues. Then, physics-based testing, such as finite element analysis can
Doppler ultrasound imaging is used to visualize local fluid flow be performed to determine the corresponding mechanical and fluid
noninvasively, where microbubbles can be used to further enhance flow properties. This provides a virtual testing tool to rapidly evaluate
its sensitivity in order to visualize blood flow in microvasculature [3]. the properties of the designed tissue construct before bioprinting it,
As conventional ultrasound imaging does not support quantitative which is a costly and time consuming endeavor.
measurement of tissue mechanical properties, ultrasound elastography As each bioprinting modality supports different process plans and
can be used to image the mechanical properties of regenerating tissues toolpath schemes, the associated compatible software system (specific
in a quantified and accurate manner [57]. to the bioprinter) allows the user to follow specific steps for a successful
In addition to abovementioned medical imaging modalities, nuclear bioprinting mission. For example, working mechanism of the majority
medicine (i.e., positron emission tomography (pet) and single, photon of extrusion-based bioprinters are similar to that of fused-deposition
emission computed tomography (SPECT)) has been used as a powerful modeling (FDM)-based 3D printers [25]. Other modalities such as
tool in order to visualize functional and molecular data of the imaged droplet-based (i.e. inkjet bioprinting) or laser-based bioprinting mod-
target [58]. Progenitor and stem cells within engineered tissue alities have different mechanisms and build platforms which require
constructs have been recently visualized in high resolution in order bioprinter-specific design software. For example, piezo-inkjet bioprin-
to monitor their metabolic and functional activities in regenerating ters (e.g., Jetlab from Microfab Inc.) have software that allows the user
cranium tissue within a critical-size defect model on rabbits [59]. to generate a code to define patterns of droplets. Laser-based bioprin-
Selecting the right imaging modality for design for bioprinting ters, such as laser-induced forward transfer-based bioprinters [68],
purposes depends on the tissue type to be fabricated as well as other though they are not commercially available, have a similar capability in
requirements such as image resolution and quality. Recently, research- allowing the operator to define different user-specific patterns.
ers have also attempted using multimodal imaging, where multiple Stereolithography and its modifications work with manufacturers’
modalities can be used in a complementary manner such as MRI/CT, software platforms or simple algorithms to define masks using
CT/PET and CT/SPECT [60]. A detailed comparison of medical image Microsoft PowerPoint slides with white and black patterns defining
modalities used in tissue engineering is presented in Table 2. transparent and opaque sections, respectively [69]. The transparent
sections allow ultraviolet light (UV) transmission through masks and
3.4. Image segmentation initiate photopolymerization.
All these software capabilities enable generation of an internal
The acquired non-invasive images (i.e., in the digital imaging and architecture with a uniform composition with limited design flexibil-
communications in medicine (DICOM) format or tagged Image file ities to the user, i.e. distance control between printed material and the
format (TIFF)) are processed through image segmentation using image dimensions of the material footprint. In order to design tissue
processing software such an open-source ITK-Snap [61] or other constructs with complex internal architecture to recapitulate native

4
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

Table 2
Comparison of medical image modalities in tissue engineering and regenerative medicine (some of data obtained from [3,46,119]).

Modality Resolution Scan time Maximum Preferred Advantages Disadvantages


volume tissue types

MRI 250×250×500 µm 5–40 min Human body Soft and hard • No risk of radiation exposure • Intolerant of in-vivo medical implants
tissue (bone) • Enables imaging of moving
components in the body in 4D
and body motions during scanning
process
such as heart contractions • Requires longer scan times for better
quality images
• Patients may experience anxiety and
claustrophobia
μMRI 5–200 µm 6–24 h Full body Soft and hard • No risk of radiation exposure • High cost
tissue (bone) • Enables incorporation of contrast • Low portability
agents for molecular imaging • Longer scan times
• Enables real time imaging
CT 0.24–0.33 mm 5 min (8–40 s
of actual scan
Human body Hard tissue
(bone)
• Tolerant of in-vivo medical • Radiation exposure limits exposure level
implants (i.e., metallic or with and frequency; not recommended for
time) battery) pregnant or breast-feeding women
• High resolution imaging of • May compromise immune system in
various tissue types children
• More tolerant of
movement during scanning
patient • Iodine contrast material can generate
allergic reaction
process
μCT 1–200 µm 2–4 h Whole rat Hard tissue
(bone)
• High resolution • Radiation exposure limits exposure level
and frequency
Ultrasound 1×1.5×0.2 mm 10–15 min Neonatal, blood All tissue types • Enables real-time images of • Image quality inferior to other modalities
vessels moving components • Cannot detect all abnormalities and
• Easy to operate and does not
involve radiation
melanomas

tissues and organs with greater fidelity, sophisticated design architec- different pore making models such as octahedron, tetrahedron, trian-
tures emerged [4], primarily in design of the porous architecture of gular and square prisms with tunable geometry as the geometric
tissue constructs. dimensions were coded as variables in the designed algorithms.
Performing Boolean operations, such as combination of pore-making
elements or structure-making primitives, Sun et al. developed a library
4.1. Computer-aided design (CAD)-based systems of unit cells using different CAD-based primitives (see Fig. 2A) [12,19].
These unit-cells can be further assembled into larger scale tissue
Computer-aided design-based systems have been used in designing constructs such as femur and spine. In addition, different unit cells
tissue constructs for bioadditive manufacturing platforms [70]. can be combined in a composite structure in order to further enrich the
Modeling approaches, such as constructive solid geometry (CSG), heterogeneity of constructs. Despite their great benefit and wide
boundary representation (B-Rep) and spatial occupancy enumeration utilization in the generation of porous tissue scaffolds, CAD-based
(SOE) [71], can be utilized to develop the design architecture of tissue systems are hampered by their poor efficiency and performance in
constructs. Constructive solid geometry modeling relies on solid generating biomimetic porous architecture or non-Euclidian solids and
primitives and Boolean operations to generate different design models; have limited control of the biomechanical properties imparted to tissue
whereas B-rep utilizes boundary elements (i.e., vertices, edges and construct as a whole [70].
faces, etc.) in order to define closed objects. Spatial occupancy
enumeration, on the other hand, combines a number of building
blocks, in general, cubic unit elements, to represent a larger scale 4.2. Image-based systems
complex solid objects. Each of these techniques has some limitations
and weaknesses with respect to each other. For example, CSG and SOE Current CAD-based modeling systems are highly time-consuming
rely on joining a number of solid primitives, which is computationally and computationally-expensive platforms for design and modeling of
expensive and requires a significant amount of data storage; on the tissue constructs. First, medical images are stacked then segmented to
other hand, B-Rep uses boundary elements enclosing the entire object. generate the surface model, which is eventually filled by generated unit
Constructive solid geometry and SOE conveniently model complex cells in a repetitive manner. The image-based design approach was first
shapes compared to B-Rep. Although these approaches are often proposed early in 2000s by Hollister et al. [73], where the defected
utilized in design of tissue constructs using commercial CAD software region of interest in medical images such as CT-scans or MRI were
such as PTC Creo, SOLIDWORKS and CATIA, current printed samples identified and filled by a stack of two 3D images made of binary unit
consist of cylindrical rasters of bioink materials, which are not easily cells. In addition to tissue constructs with regular porous architecture,
approximated using these modeling approaches. Therefore, researchers irregular porous scheme can be generated by assigning a random
used pore making elements from a design library and subtracted them number generator-based approach, where different voxel states are
from the external surface geometry in order to create porous archi- generated randomly throughout the defect [74]. The proposed ap-
tecture. With this, an internal architecture with controlled porosity proach was first implemented for defects in mandibular bone (see
(i.e., pore size, level and geometry) can be obtained. For example, Fig. 2B1–B4) and orbital floor segments, and sample tissue constructs
researchers created algorithms that enabled automatic subtraction of were 3D printed using a selective-laser sintering process [75].
the negative geometry from a CAD model of a femur bone segment With direct manipulation of the 3D design architecture using the
obtained from CT images [72]. By subtracting the negative geometry abovementioned image-based approach instead of generating CAD
(that are assembled in 3D as periodic elements) of the pore-making models, defect models were rapidly generated, which also provided
structure from the surface model of the femur bone segment, 3D additional freedom to the user in imposing variations in the porosity
porous lattice structures were obtained. The authors also presented network by assigning different design architectures to discrete sections

5
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

Fig. 2. Blueprint modeling used in design of tissue constructs: (A) A library of unit cells constructed from different primitives using CAD-based systems (reproduced and adapted with
permission from [12]). Image-based design of mandibular condyle scaffolds, where (B1) CT scan of Yucatan minipig condyle was used to generate the external anatomical shape. (B2) A
database of porous architecture was used to generate internal architecture, which was then combined with the (B3) external geometry in order to generate (B4) the complete construct
(reproduced and adapted with permission from [75]). Freeform design of a wound device, where (C1) the wound bed surface was captured using image intensities and (C2) a NURBS
surface was fitted followed by (C3) partitioning the geometric domain using lofting process (reproduced and adapted with permission from [77]). Triply-periodic minimal implicit
surfaces including cells of (D1) C(D), (D2) Gyroid and (D3) Manta 35 (reproduced and adapted with permission from [84]. Space filling curves (E1) using 0/45/90/135° degrees of
filament orientation in sequential layers, (E2-E3) Hilbert recursive curves in two consecutive layers (reproduced and adapted with permission from [101]).
6
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

within the defect site. into two intertwined labyrinthine domains that are periodic in three
distinct lattice directions [85]. Using TPMSs, Schwarz's Diamond and
Schoen's Gyroid designs were 3D printed for tissue biofabrication
4.3. Freeform systems
[86,87]. Heterogeneous porous scaffold architectures were also demon-
strated in a linearly-graded way to explore the mechanical responses
Computer-aided design and image-based systems provide porous
for stretching and bending dominated deformations for bone tissue
architectures; however, they mostly generate design architectures with
engineering [88].
uniform porosity. Freeform-based systems; on the other hand, generate
a tissue construct geometry with variations in its architecture since the
geometric domain can be partitioned into different regions [76]. These 4.5. Space filling curves
regions can then be populated by different materials and porous
architectures. In this regard, wound devices were developed using a Computer-aided design- or image-based systems utilize solid
surface model obtained from 2D wound images; which were then primitives as building blocks for larger-scale tissue construct models,
projected into 3D using image intensity function in Image J software which are difficult to support with most 3D bioprinting techniques.
(see Fig. 2C1) [77]. Then, the generated surface was approximated into Extrusion-based bioprinting is the most common bioprinting modality
a non-uniform rational B-spline (NURBS) surface followed by lofting in tissue fabrication, relying on the deposition of straight rasters in
process with respect to the center point of the top surface of the wound each layer using 0/90° lay-down pattern [89]. As a result, the above-
(see Fig. 2C2). Lofting, also known as skinning, is a process forming a mentioned approaches, except the freeform design approach, are
surface that is a blend among a set of curves called generators [78,79]. highly challenging to apply in EBB. Thus, an alternative approach in
This way, a number of regions were defined and each region was then designing the internal architecture of the tissue construct utilizes
filled by rasters of different materials; the authors 3D bioprinted space-filling curves as a toolpath. In the literature, design models with
several wound devices with variable biomolecule and sodium alginate different space-filling curves, such as lay-down patterns with various
concentrations [80] (see Fig. 2C3), which provided synchronous raster deposition angles (i.e., 0/45/90/135°) and Hilbert and
delivery of biomolecules with the wound healing process [76]. A similar Sierpinsky curves, have been demonstrated by Li et al. [90] and
approach was also applied to partition 3D models of vertebrae, femur Starly et al. [91]. The design architecture of the filling curves was
and aorta using mesh-based offsetting technique to control the porous changed to generate periodic sub-patterns enabling construction of
architecture [15]. each layer using EBB. For subsequent layers, the orientation of the
Using the freeform-design approach, tissue constructs can be filling curves was changed to facilitate sufficient mechanical support for
designed and bioprinted with spatial control of geometric architecture structurally-integrated constructs. Such examples have been further
and material composition as well as spatiotemporal control of release investigated using Hilbert curves through partitioning the entire design
kinetics of biomaterials and biologically active components. Recently, region into unit cells and assigning Hilbert curves at different scales at
researchers also identified means to decrease total fabrication time on each region. By combining each Hilbert curve at each region, a
freeform objects by optimizing the build direction [81]. These features continuous toolpath plan could be executed. The use of space filling
are not quite feasible using other bioprinting modeling approaches. curves provides a compatible design platform for EBB and its sub-
modalities (i.e., pneumatic, mechanical, and solenoid micro-extrusion
as detailed Ozbolat et al. [25]), which is not quite possible using other
4.4. Design using implicit surfaces
bioprinting techniques. A similar approach was also performed based
on a Lindenmayer system in order to biomimic vascular networks,
One of the major drawbacks of CAD- and image-based systems is
where branched structures were printed using LBB [92].
the inability to present highly complex shapes in a short period of time.
In summary, a number of blueprint modeling approaches have been
The geometric complexity of the design is limited by the degree of
utilized in design for bioprinting and each of them possesses different
complexity of the primitives. Thus, implicit functions have been used to
capabilities. Table 3 compares different blueprint modeling techniques
illustrate periodic minimal surfaces for complex-shape tissue scaffolds,
and presents their strengths and limitations along with their compat-
which can accurately control the spatial porosity distribution within an
ibility with different bioprinting modalities.
arbitrarily shaped architecture [82,83]. Minimal surfaces have several
advantages such as possessing lightweight architecture and demon-
strating high structural strength [84]. In this regard, triply periodic 5. Toolpath planning for bioprinting
minimum surfaces (TPMSs) have been established as shown in
Fig. 2D1–D3. TPMSs are smooth infinite surfaces that split the space There are currently two different toolpath planning approaches

Table 3
Comparison of blueprint modeling techniques used in tissue construct bioprinting.

Blueprint Advantages Disadvantages Compatible bioprinting References


modeling modalities
techniques

CAD-based systems Easy to model; accurately represent simple primitive Computationally expensive; the DBB and LBB processes based [19,72]
shapes complexity is limited to the unit cell on cell transfer
resolution; slow processing
Image-based systems Directly uses medical images and eliminates their 3D Low accuracy; is not compatible with DBB LBB processes based on [16,73,75]
segmentation; facilitates anatomical architecture; and LBB photopolymerization
Freeform systems Computationally efficient; facilitates heterogeneous Not compatible with majority of CAD DBB, LBB and EBB [76,77,80]
design environment; supports geometric complexity; software in the market
Implicit surfaces Computationally efficient; facilitates heterogeneous Not compatible with majority of CAD LBB [84,86,87]
design environment; support geometric complexity; software in the market
generates structurally strong design models
Space filling curves Computationally efficient; facilitates heterogeneous Does not support geometrically complex EBB [91,96]
design environment; supports most of 3D bioprinters models
and FDM-based 3D printers

7
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

used in bioprinting of tissue and organ constructs. Bioprinting of bulky higher feed rate while approaching sharp turns; however, it is not
tissue constructs do not require highly specialized toolpath planning as completely avoidable. A rational approach to such a problem is to
the goal is to fill a simple primitive geometry such as cubic tissue generate a tangential continuity (C1) along the toolpath.
constructs. Therefore, standard toolpath plans can easily fill the Toolpath planning in Cartesian form can be also problematic if
geometry without the need for precision. Such examples are bioprint- functionally-graded tissue models are desired. For example, a toolpath
ing of non-porous bulky constructs such as tissue constructs [93] and plan in Cartesian form was used to build multi-functional wound
tissue models [8,94] for regenerative medicine and pharmaceutics, devices (see Fig. 3B1–B2) [77]. A 3D blending process was developed
respectively. to generate functionally-graded wound devices with different bioma-
In the literature, two different toolpath planning approaches have terial and loaded biomolecule concentrations in different regions. The
been proposed including bioprinting rasters in (i) Cartesian and (ii) developed toolpath plan in Cartesian coordinates is, however, no more
parametric form. As the use of DBB and LBB is greatly less than that of than an approximation of the actual gradient since the blending
EBB in the bioprinting community and bioprinters associated with direction is independent of toolpath deposition direction. As shown
these bioprinting modalities utilize manufacturer specific toolpath in Fig. 3C1, the toolpath follows Cartesian coordinates, but does not
planning strategies (in Cartesian form), the author here exclude follow blending directions between two features. Thus, the toolpath
toolpath planning in DBB and LBB modalities. does not capture the geometry; smaller raster size with smaller gaps
between adjacent rasters is required for improved accuracy. Moreover,
5.1. Toolpath planning in Cartesian form jumps or motion without deposition is substantial due to the nature of
the toolpath that is independent of feature geometry.
Toolpath planning in Cartesian (x, y, z) form has been widely used When any hollow features are introduced into a tissue construct,
in all bioprinting modalities including EBB, DBB and LBB due to its such as lumen in a blood vessel, toolpath planning in Cartesian
simplicity in creating layers of rasters stacked up in 3D. As EBB has coordinates brings further problems. Almost all of the toolpath
been the most popular among all bioprinting modalities, toolpath planning in the literature proposes similar deposition pattern in
planning in EBB has attracted researchers [77]. At each layer during Cartesian coordinates when any hollow shapes are imposed in object
EBB, the deposition direction is reoriented to create a 0/90° lay-down architecture [96]. Qiu et al. [97] addressed this problem; however, the
pattern. Although different angles can be used (i.e., 0/45° or 0/135°), developed methodology amounted to little more than an improvement
0/90° allows for better mechanical and structural integrity than other in space filling. In that work, they minimized the gap by introducing an
angles, thus it is commonly used in bioprinting as well as FDM-based adaptive roadwidth toolpath generation method. But the introduction
3D printing techniques [95]. For example, Fig. 3A1–A2 a toolpath of several hollow shapes resulted in too many starts and stops during
designed for continuous deposition and minimal fluid congestion; deposition. The deposition process again followed Cartesian coordi-
however, a problem with excess accumulation of bioink at the point nates independent of hollow feature geometry. Therefore, parametric
of directional change exists with the zigzag pattern. When approaching modeling of the geometric domain that is compatible with the toolpath
sharp turns, deposition stops, the direction changes and the feed rate plan can be a solution to such a problem. An example is provided in
accelerates in the new direction. Although continuous deposition is Fig. 3C2, where the toolpath plan follows the directional gradient in
observable, there is considerable degeneration in the uniformity of bioink composition.
deposited filament thickness. This can be alleviated by selecting a

Fig. 3. Toolpath planning in Cartesian coordinates: (A1) design of a continuous toolpath with variable porosity and (A2) bi-layer bioprinted vertebrae; (B1) design of a toolpath for
functionally-graded wound devices with variable bioink composition and (B2) bi-layer bioprinted device; comparison of toolpath planning for functionally-graded tissue constructs
designed using (C1) Cartesian and (C2) parametric coordinates.

8
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

Fig. 4. Toolpath planning in parametric coordinates. (A1) Medical images were processed (A2) in order to generate a CAD model of aorta, where (A3–A4) a toolpath plan was developed
with controlled material composition along the parametric distance u (different colors represent different materials, reproduced and adapted with permission from [96]). (B1) Using
medical image processing, (B2) a femur model was generated and sliced (B3) for toolpath planning with controlled porosity increasing the parametric distance u. (B4) A sample double-
layer structure was bioprinted using sodium alginate hydrogel (reproduced and adapted with permission from [5]).

5.2. Toolpath planning in parametric form ing modalities enable a resolution at the submicron level but smaller
features that govern cell attachment, guidance and spreading cannot be
In order to overcome the issues in toolpath planning with Cartesian achieved easily. Therefore, Mironov et al. devised hybrid fabrication
coordinates, toolpath planning in parametric (u, v) coordinates has techniques such as bioprinting cells or cell aggregates onto electrospun
been of interest to researchers. In this regard, Ozbolat et al. proposed fibers [100]. Nevertheless, little design consideration has been given to
the idea of using parametric ruling lines in toolpath planning for multi-scale modeling of tissue constructs having both nano- and micro-
complex geometric domains, which can dictate the material and scale composition. As multi-scale modeling of tissue constructs may be
porosity composition (see Fig. 4) [10,14,96,98]. First, a medical a computationally expensive process, the modeling can be done using
image-generated CAD file was sliced into several layers and for each computationally-efficient approaches such as implicit surfaces in multi-
layer, NURBS splines were fitted into the curves. Next, the curves were ple scale.
parameterized and sampled into points, and parametric ruling lines The majority of research attempts in design and fabrication of
were generated between NURBS splines. Then, toolpath generation tissue constructs have been performed based on uniform composition
algorithms were developed to create the toolpath plan in two con- of design and fabrication elements [101]. For example, bioprinted
secutive layers. The first layer was based on ruling lines and followed a tissue constructs utilizing uniform micro-architectural and material
zigzag approach, where arc fitting was introduced at the turns. Arc composition throughout their geometry, but native tissues are anato-
fitting was used to eliminate two consecutive sharp turns in the zigzag mically heterogeneous. Therefore, new design approaches should be
pattern and enabled C1 tangent continuity with constant feedrate of developed in order to achieve heterogeneity in tissue constructs with
deposition. This eliminates the aforementioned shortcomings of tradi- functionally-graded or hierarchical properties. Unfortunately, such
tional and recent approaches to hollow object prototyping. In order to models with controlled porosity, material composition and biomolecule
support the zigzag pattern along the ruling lines, a spiral toolpath plan distribution have rarely been considered in design for bioprinting [4].
was proposed for the next layer, which was closely perpendicular to Significant advances have been made in the design of tissue
ruling lines at the junction point providing better structural support for constructs with intricate geometries in 3D as the design process is
the next layer. The advantages of the spiral toolpath are that it (1) quite flexible with minimal limitations. However, manufacturing
enables a continuous deposition through a hollow feature compared to techniques that incorporate 3D printing processes set stricter limita-
the conventional zigzag-based space-filling technique and (2) a smooth tions as highly intricate shapes can only be manufactured using micro-
transition from one material composition property to another. stereolithography and two-photon polymerization (2PP) processes.
Although these techniques facilitate 3D printing of high-definition
6. Limitations tissue scaffolds, cells have not yet been bioprinted using these
techniques. Particularly, 2PP polymerization enables printing of com-
Despite the significant progress in designing scaffolds and tissue plex geometries such as porous lattice structures with a resolution of
constructs as generic models, each tissue and organ type has unique less than a few hundred nanometers [102]. Although this technology
requirements and characteristics [12]. Therefore, behavior of different may generate structures too small to accommodate high cell densities,
cell types may vary depending on the design architecture of the it can be used to generate hybrid scaffolds where cells can be bioprinted
constructs. For example, the pore size and porous architecture needs on top of prefabricated scaffolds.
to be different for specific cell types. For cells that are metabolically Use of soft materials do not support the original design model. The
highly active such as cardiac or pancreas cells, media transport is majority of bioprinting work relies on the use of soft materials such as
crucial; therefore, high porosity for effective transport is of major hydrogels and cell aggregates, and these materials do not retain their
importance. But for cells such as chondrocytes, the normal cellular original shape after bioprinting due to swelling, dehydration, contrac-
environment is hypoxic, thus a less porous architecture would better tion, spreading and splashing of the bioink materials. Fig. 5A1–A2
support formation of cartilage [99]. demonstrate 3D printed heart valve constructs made of poly(ethylene
A prominent weakness of current bioprinting modalities is the lack glycol)-diacrylate (PEG-DA) at different scales, where the shape fidelity
of precise control on the cellular micro-environment. Current bioprint- analysis showed that the deviation from the design model increased

9
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

Fig. 5. Recent approaches in bioprinting of high fidelity tissue constructs. (A1) 3D printed heart valve constructs fabricated at different scales using EBB with a photo crosslinking
apparatus and (A2) their fidelity analysis showing deviations from the original design (reproduced and adapted with permission from [120]). (B) Step-by-step bioprinting trajectory with
predictive compensation in order to inkjet bioprint vascular tubular constructs within an acceptable cylindricity tolerance (reproduced and adapted with permission from [105]. (C)
Step-by-step bioplotting within a hydrogel bath loaded with gelatin microparticles enabling the retention of the original shape of the deposited bioink, where the bioplotted construct was
removed when the hydrogel bath was liquefied at 37 °C (reproduced and adapted from [107])). Scale bars correspond to 1 cm.

when the size of the tissue construct diminished. In DBB [26] and some the designed toolpath, toolpath compensation approaches can be
of LBB processes [31], similar issues can be observed as well even if developed and integrated according to the behavior of different bioink
these processes facilitate bioprinting in higher resolution. Depending materials. Such toolpath compensation has been widely used in
on the process parameters, crosslinking and rheological properties of computer numeric control (CNC) machining and other traditional
the bioink solution, the shape accuracy may deviate from the design manufacturing techniques [103]. For example, Huang et al. demon-
models. The majority of current toolpath planning approaches do not strated piezo-inkjet bioprinting of tubular vascular constructs by
account for the changes in the shape of soft bioink materials due to depositing sodium alginate into a crosslinker pool by utilizing the
reasons discussed previously. As such changes result in variations from buoyancy force in order to support overhangs [104]. Due to the

10
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

substantial weight of the bioink compared to the buoyancy force, gels, where cells are embedded in hydrated scaffolds and have the
bioprinted constructs were deformed and significantly deviated from ability to migrate in a 3D space [109]. In contrast, cells in traditional
the original design during construction. Therefore, the group later polymeric scaffolds are limited to migrate on the scaffold surface only.
developed a toolpath strategy with predictive compensation regarding The porosity within the hydrogel network thus plays a crucial role in
the deflection on bioprinted tubular constructs, which in turn resulted stimulating cell growth and proliferation. Although there is a great
in accurate bioprinting of the design model as can be seen in Fig. 5B need for micro-porous design architecture, such a design has not yet
[105]. In EBB, similar issues are also observed for the majority of been attempted.
hydrogels used as a bioink material. Although compensation in the In addition to cell-laden hydrogels, design consideration should
lateral plane is important in order to accurately extrude the bioink in also be given to scaffold-free bioprinting. As there is no scaffold to
the proper position; compensation in z-axis is also important as model and design, attention should focus on design of building blocks
majority of the hydrogels contract in z-axis during bioprinting. such as tissue spheroids [110] and tissue strands [111] or other form of
Additionally, the distance between the nozzle tip and the printing aggregates. As cell aggregates are subjected to contraction due to
plane is prone to increase over time resulting in imperfections or even cadherin-induced cell-cell integration, the impact of aggregate shape
failures. Thus, the distance from the nozzle tip to the bioprinting stage change on the overall tissue fabrication process demands careful
should be adjusted automatically during the bioprinting process and consideration [112]. There has been a significant amount of theoretical
should dynamically inform the toolpath plan as such toolpath was work in modeling fusion dynamics of the 3D constructs but the
demonstrated by Ozbolat et al. [106]. In order to overcome the majority of these studies do not inform the initial tissue design in a
limitations of bioprinting soft bioink materials, Feinberg's group systematic way. Most important, there is no design-driven system that
recently demonstrated a bioplotting process, where the bioink material is well-integrated with the bioprinting process. For example, a recent
was extruded and plotted into a hydrogel bath with gelatin micro- work demonstrated toolpath planning for scaffold-free bioprinting of
particles enabling the bath behaving like a Bingham plastic [107]. In vascular tissue construct [113]. The authors demonstrated an extensive
other words, at 22 °C, the bath acts like a liquid facilitating extrusion work on generating the toolpath plan for bioprinting including the path
and plotting of the bioink when the nozzle moves under shear stress plan for the support material (agarose) and the main material (cell
and acts like a solid holding the plotted bioink in place when the pellet). Upon culturing the samples for sufficient time, cells aggregated
influence of the nozzle-induced shear stress diminishes. Using this and the support material was removed. The design toolpath and the
approach, high fidelity constructs were plotted into the hydrogel bath fabricated structure differ considerably from each other, and such
and easily removed at 37 °C when the hydrogel bath liquefied (see compensation should be incorporated into the design. Thus, exploring
Fig. 5C). the underlying physics of the process and informing the toolpath
Despite the widespread use of the presented blueprint modeling in decision based on such knowledge will be valuable in optimization of
design of tissue constructs, a limited number of them are directly the toolpath plan and fabrication of the tissue constructs, respectively.
relevant and compatible with bioprinting processes. Extrusion-based There is now a current trend in scaffold-free fabrication of tissues and
bioprinting is compatible with space filling curves as the bioprinter physics-based model informed design will enable accurate fabrication
head and hence, the deposition can directly follow the space filling of tissues.
curves in a continuous manner. Other blueprint modeling approaches Physics-based design of cell-laden hydrogels that account for
can also be used in bioprinting; however, the accuracy of the design structural and biological changes over time will be highly valuable to
model in general tends to vary greatly from the bioprinted physical the bioprinting research community. Although modeling of degrada-
models. Conversely, DBB techniques are more compatible with CAD- tion and media diffusion has been done in polymeric porous scaffolds
based models as droplets of spherical shape are deposited as building [114], degradation of matrix material as well as proliferation of cells in
blocks. Therefore, the sphere can be used as a “primitive” to create the cell-laden hydrogels has not been investigated thoroughly. A thorough
blueprint models for droplet-based bioprinting techniques. For LBB, understanding of long-term structural stability of hydrogels is of key
submodalities determine the choice of the appropriate blueprinting importance as most of the hydrogels undergo a rapid degradation
technique. For example, LBB processes involving photopolymerization process. For example, a hydrogel-based environment has been used in
are compatible with majority of blueprint techniques except the space the fabrication of microfluidic devices to understand cancer metastasis
filling curves. As such processes have high-resolution capabilities, they [115]. But, hydrogel materials such as fibrin have quick gelation rates
are able to capture the blueprint models accurately. Laser-based and the device did not retain its integrity for more than two weeks.
bioprinting processes based on cell transfer have requirements similar Researchers have blended different gels such as hyaluronic acid or
to DBB. As the bioink is deposited on the substrate, it forms a droplet; collagen to improve the degradation and mechanical properties [116].
thus, a CAD-based system with spherical primitives is highly compa- However, experimentation with combinations of materials at different
tible. concentrations is an expensive and time consuming endeavor. Thus,
Despite the great number of off-the-shelve 3D printers in the multi-physics based design models at multiple-scale that can predict
market, very few are open-source printers that allow the user to define the degradation profile of the matrix environment, and the temporal
any toolpath plan on demand using G-code [108]. Such a capability and structural impact of cell proliferation and extracellular matrix
would enable the operator to load a CAD model or write customized (ECM) deposition on the tissue construct will be an invaluable tool in
codes according to design needs; however, non-open-source 3D the development of novel biomaterials and bioprinting techniques.
printers do not allow such flexibility and the operator is limited to Biologically informed design represents a promising approach in
software-specific automatic toolpath algorithms used for the input CAD the design and biofabrication of next generation tissue constructs.
model. This feature is particularly useful for researchers who bioprint Natural tissues possess a heterogeneous architecture with different cell
simple primitive shapes in experimental settings where the added time and ECM compositions throughout their structure; this phenomenon
expenditure of designing a custom toolpath is not required. Rather, the should be incorporated into the design environment. Currently, these
operator can quickly generate the toolpath directly. variations are addressed in a rudimentary fashion rather than devising
sophisticated design concepts. Moreover, physics-based modeling of
7. Future directions various parameters such as tissue formation and degradation of
scaffold-matrix will be highly informative and can contribute signifi-
The majority of the design considerations in tissue engineering have cantly toward a complex design environment. Considerations in
been given to scaffold design for cell attachment and growth; however, current design methods for bioprinting of tissue constructs are limited
the current trend in biofabrication is moving toward cell-laden hydro- to optimization of porosity for sufficient fluid flow and maximization of

11
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

mechanical and structural integrity. Although porosity is important in engineering applications, Tissue Eng. Part B Rev. 21 (2014) 1–44.
[4] I.T. Ozbolat, A. Khoda, M. Marchany, J.A. Gardella, B. Koc, Hybrid tissue scaffolds
the tissue regeneration process, several other factors affect tissue for controlled release applications: a study on design and fabrication of hybrid and
regeneration such as cell-cell and cell-matrix interactions, scaffold heterogeneous tissue scaffolds for controlled release applications is presented in
degradation, cell microenvironment in differentiating and proliferating this paper, Virtual Phys. Prototyp. 7 (2012) 37–47.
[5] A.K.M. Khoda, I.T. Ozbolat, B. Koc, Designing heterogeneous porous tissue
cells and vascularization; however, majority of these factors have scaffolds for additive manufacturing processes, CAD Comput. Aided Des. 45
largely been ignored in currently available design approaches and (2013) 1507–1523.
tools. Therefore, more comprehensive design systems need to be [6] H. Janik, M. Marzec, A review: fabrication of porous polyurethane scaffolds,
Mater. Sci. Eng. C 48 (2015) 586–591.
developed in order to integrate other crucial biological factors into [7] A.G. Mikos, J.S. Temenoff, Formation of highly porous biodegradable scaffolds for
the process of building successful tissue constructs. tissue engineering, Electron. J. Biotechnol. 3 (2000) 114–119.
The majority of the design software currently offered by the [8] W. Peng, D. Unutmaz, I.T. Ozbolat, Bioprinting towards physiologically relevant
tissue models for pharmaceutics, Trends Biotechnol. 34 (2016) 722–732.
bioprinter companies, are limited to predefined toolpath planning
[9] U.A. Gurkan, R. El Assal, S.E. Yildiz, Y. Sung, A.J. Trachtenberg, W.P. Kuo, et al.,
strategies. For example, the design software enables the user to input Engineering anisotropic biomimetic fibrocartilage microenvironment by bio-
a CAD model (that can be obtained from medical images such as CT printing mesenchymal stem cells in nanoliter gel droplets, Mol. Pharm. 11 (2014)
scan and MRI) and fill contours of each slice generated by the design 2151–2159.
[10] A.K.M. Khoda, I.T. Ozbolat, B. Koc, A functionally gradient variational porosity
model similar to FDM-based additive manufacturing processes. But architecture for hollowed scaffolds fabrication, Biofabrication 3 (2011) 034106..
each tissue type or use of different bioink materials and corresponding [11] Z. Fang, B. Starly, W. Sun, Computer-aided characterization for effective
crosslinking mechanisms necessitates a different path planning strat- mechanical properties of porous tissue scaffolds, CAD Comput. Aided Des. 37
(2005) 65–72.
egy. For instance, Pluronic gel has been used as a support material in [12] W. Sun, B. Starly, A. Darling, C. Gomez, Computer-aided tissue engineering:
bioprinting processes as it can quickly retain its shape upon bioprinting application to biomimetic modelling and design of tissue scaffolds, Biotechnol.
[117]. This process can be fine-tuned by using the appropriate Appl. Biochem. 39 (2004) 49–58.
[13] A.K.M.Khoda, I.T.Ozbolat, B.Koc, Functionally gradient tissue scaffold design and
concentration of the gel and its temperature profile; however, the deposition path planning for bio-additive processes, in: Proceedings of the
other gels utilized in bioprinting do not allow for such a well-defined Industrial and Systems Engineering Research Conference, San Juan Puerto Rico,
retention of the gel shape post-bioprinting. Therefore, the design 2013.
[14] A.K.M. Khoda, I.T. Ozbolat, B. Koc, Modeling of variational gradient porous
models do not actually account for the changes in the printed hydrogel
architecture with multi-directional filament deposition in 3D scaffolds, Comput.
geometry nor compensate for changes in shape caused by swelling and Aided Des. Appl. 10 (2013) 445–459.
dehydration. Therefore, future design software should take into [15] A. Khoda, I.T. Ozbolat, B. Koc, Engineered tissue scaffolds with variational porous
architecture, J. Biomech. Eng. 133 (2011).
account bioink solution properties and should enable the user to define
[16] S.J. Hollister, R.D. Maddox, J.M. Taboas, Optimal design and fabrication of
distinct path plans for different hydrogels of varying concentrations. scaffolds to mimic tissue properties and satisfy biological constraints, Biomaterials
Many researchers entering into the bioprinting field invest substantial 23 (2002) 4095–4103..
effort into designing custom toolpath plans for hydrogels. Currently [17] S.J. Hollister, C.Y. Lin, Computational design of tissue engineering scaffolds,
Comput. Methods Appl. Mech. Eng. 196 (2007) 2991–2998.
available software does not meet the demands for more sophisticated [18] C.Y. Lin, N. Kikuchi, S.J. Hollister, A novel method for biomaterial scaffold
bioprinting constructs. internal architecture design to match bone elastic properties with desired porosity,
J. Biomech. 37 (2004) 623–636.
[19] M.A. Wettergreen, B.S. Bucklen, B. Starly, E. Yuksel, W. Sun, M.A.K. Liebschner,
8. Conclusion Creation of a unit block library of architectures for use in assembled scaffold
engineering, CAD Comput. Aided Des. 37 (2005) 1141–1149.
This paper elucidates design requirements for bioprinting of tissue [20] M.A. Wettergreen, B.S. Bucklen, W. Sun, M.A.K. Liebschner, Computer-aided
tissue engineering of a human vertebral body, Ann. Biomed. Eng. 33 (2005)
and organ constructs, and presents the design process for bioprinting 1333–1343.
starting with acquisition of patient-specific images using medical image [21] Y. Zhang, Y. Yu, I.T. Ozbolat, Direct bioprinting of vessel-like tubular microfluidic
modalities (including CT and MRI), followed by blueprint modeling channels, J. Nanotechnol. Eng. Med. 4 (2013) 020902.
[22] Y. Zhang, Y. Yu, A. Akkouch, A. Dababneh, F. Dolati, I.T. Ozbolat, In vitro study of
and a critique of currently available toolpath planning approaches for directly bioprinted perfusable vasculature conduits, Biomater. Sci. 3 (2015) (134–
3D bioprinting purposes. Contemporary blueprint modeling ap- 13).
proaches, used in creating internal architecture of porous tissue [23] F. Dolati, Y. Yu, Y. Zhang, A.M. De Jesus, E.A. Sander, I.T. Ozbolat, In vitro
evaluation of carbon-nanotube-reinforced bioprintable vascular conduits,
constructs, are detailed and discussed along with strengths and
Nanotechnology 25 (2014) 145101.
limitations. As native tissues and organs are heterogeneous in nature, [24] V.K. Lee, D.Y. Kim, H. Ngo, Y. Lee, L. Seo, S.-S. Yoo, et al., Creating perfused
bioprinting processes should be informed by the appropriate blueprint functional vascular channels using 3D bio-printing technology, Biomaterials 35
modeling techniques in order to recapitulate the native form and (2014) 8092–8102.
[25] I.T. Ozbolat, M. Hospodiuk, Current advances and future perspectives in extru-
function, and bioprinting should complete the process using an sion-based bioprinting, Biomaterials 76 (2016) 321–343.
appropriate compatible toolpath planning approach. [26] H. Gudapati, M. Dey, I. Ozbolat, A comprehensive review on droplet-based
bioprinting: past, present and future, Biomaterials 102 (2016) 20–42.
[27] A. Ovsianikov, M. Gruene, M. Pflaum, L. Koch, F. Maiorana, M. Wilhelmi, et al.,
Acknowledgements Laser printing of cells into 3D scaffolds, Biofabrication 2 (2010) 14104.
[28] I.T. Ozbolat, Y. Yu, Bioprinting toward organ fabrication: challenges and future
This work has been supported by National Science Foundation trends, IEEE Trans. Biomed. Eng. 60 (2013) 691–699.
[29] K. Arcaute, B.K. Mann, R.B. Wicker, Stereolithography of three-dimensional
CMMI Awards 1349716 and 1462232, Diabetes in Action Research and bioactive poly(ethylene glycol) constructs with encapsulated cells, Ann. Biomed.
Education Foundation Grant 426, and Osteology Foundation Grant Eng. 34 (2006) 1429–1441.
15–042. The author would like to thank Donna Sosnoski from Ozbolat [30] A. Ovsianikov, S. Mühleder, J. Torgersen, Z. Li, X.-H. Qin, S. Van Vlierberghe,
et al., laser photofabrication of cell-containing hydrogel constructs, Langmuir 30
Lab at Penn State University for her assistance with the typesetting. (2014) 3787–3794.
The author confirms that there are no known conflicts of interest [31] B.C. Riggs, a.D. Dias, N.R. Schiele, R. Cristescu, Y. Huang, D.T. Corr, et al.,
associated with this publication and there has been no significant Matrix-assisted pulsed laser methods for biofabrication, MRS Bull. 36 (2011)
1043–1050.
financial support for this work that could have influenced its outcome.
[32] J.-H. Park, R. a. Pérez, G.-Z. Jin, S.-J. Choi, H.-W. Kim, I.B. Wall, Microcarriers
designed for cell culture and tissue engineering of bone, Tissue Eng. Part B Rev.
References 19 (2013) 172–190.
[33] L. Koch, A. Deiwick, S. Schlie, S. Michael, M. Gruene, V. Coger, et al., Skin tissue
generation by laser cell printing, Biotechnol. Bioeng. 109 (2012) 1855–1863.
[1] R. Langer, J.P. Vacanti, Tissue engineering, Science 260 (1993) 920–926. [34] V. Lee, G. Singh, J.P. Trasatti, C. Bjornsson, X. Xu, T.N. Tran, et al., Design and
[2] I.T. Ozbolat, W. Peng, V. Ozbolat, Application areas of 3D bioprinting, Drug Fabrication of Human Skin by Three-Dimensional Bioprinting, Tissue Eng. Part C
Discov. Today 21 (2016) 1257–1271. Methods 20 (2013) 473–484.
[3] S.Y. Nam, L.M. Ricles, L.J. Suggs, S.Y. Emelianov, Imaging strategies for tissue [35] J. Yan, Y. Huang, D.B. Chrisey, Laser-assisted printing of alginate long tubes and

12
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

annular constructs, Biofabrication 5 (2013) 015002. Humana Press, Totowa, NJ, 2012, pp. 289–302.
[36] T. Xu, H. Kincaid, A. Atala, J.J. Yoo, High-throughput production of single-cell [70] S.M. Giannitelli, D. Accoto, M. Trombetta, A. Rainer, Current trends in the design
microparticles using an inkjet printing technology, J. Manuf. Sci. Eng. 130 (2008) of scaffolds for computer-aided tissue engineering, Acta Biomater. 10 (2014)
021017. 580–594..
[37] K. Pataky, T. Braschler, A. Negro, P. Renaud, M.P. Lutolf, J. Brugger, Microdrop [71] A.G. Requicha, Representations for rigid solids: theory, methods, and systems,
printing of hydrogel bioinks into 3D tissue-like geometries, Adv. Mater. 24 (2012) ACM Comput. Surv. 12 (1980) 437–464.
391–396. [72] C. Cheah, C. Chua, K. Leong, C.-H. Cheong, M.-W. Naing, Automatic algorithm for
[38] C. Xu, W. Chai, Y. Huang, R.R. Markwald, Scaffoldfree inkjet printing of generating complex polyhedral, Tissue Eng. 10 (2004) 595–610.
threedimensional zigzag cellular tubes, Biotechnol. Bioeng. 109 (2012) [73] S.J. Hollister, R.A. Levy, T.-M. Chu, J.W. Halloran, S.E. Feinberg, An image-based
3152–3160. approach for designing and manufacturing craniofacial scaffolds, Int. J. Oral
[39] A. Blandino, M. Macías, D. Cantero, Formation of calcium alginate gel capsules: Maxillofac. Surg. 29 (2000) 67–71.
influence of sodium alginate and CaCl2 concentration on gelation kinetics, J. [74] S.J. Hollister, Porous scaffold design for tissue engineering, Nat. Mater. 4 (2005)
Biosci. Bioeng. 88 (1999) 686–689. 518–524.
[40] K. Christensen, C. Xu, W. Chai, Z. Zhang, J. Fu, Y. Huang, Freeform inkjet [75] M.H. Smith, C.L. Flanagan, J.M. Kemppainen, J.A. Sack, H. Chung, S. Das, et al.,
printing of cellular structures with bifurcations, Biotechnol. Bioeng. 112 (2015) Computed tomography-based tissue-engineered scaffolds in craniomaxillofacial
1047–1055. surgery, Int. J. Med. Robot. Comput. Assist. Surg. 3 (2007) 207–216.
[41] J.Y. Park, J.-C. Choi, J.-H. Shim, J.-S. Lee, H. Park, S.W. Kim, et al., A [76] I.T. Ozbolat, B. Koc, Multi-function based modeling of 3D heterogeneous wound
comparative study on collagen type I and hyaluronic acid dependent cell behavior scaffolds for improved wound healing, Comput. Aided Des. Appl. 8 (2011) 43–57.
for osteochondral tissue bioprinting, Biofabrication 6 (2014) 035004. [77] I.T. Ozbolat, B. Koc, 3D hybrid wound devices for spatiotemporally controlled
[42] M.F. Coakley, D.E. Hurt, N. Weber, M. Mtingwa, E.C. Fincher, V. Alekseyev, et al., release kinetics, Comput. Methods Prog. Biomed. 108 (2012) 922–931.
The NIH 3D Print exchange: a Public resource for bioscientific and biomedical 3D [78] I.T. Ozbolat, M. Marchany, J.A. Gardella Jr, F.V. Bright, A.N. Cartwright, R. Hard,
Prints, 3D Print. Addit. Manuf. 1 (2014) 137–140. et al., Feature-based design of bio-degradable micro-patterned structures,
[43] H. Xu, S.F. Othman, R.L. Magin, Monitoring tissue engineering using magnetic Comput. Aided Des. Appl. 6 (2009) 661–671.
resonance imaging, J. Biosci. Bioeng. 106 (2008) 515–527. [79] I.T. Ozbolat, B. Koc, Multi-directional blending for heterogeneous objects,
[44] T.A. Bley, O. Wieben, C.J. François, J.H. Brittain, S.B. Reeder, Fat and water Comput. Des. 43 (2011) 863–875.
magnetic resonance imaging, J. Magn. Reson. Imaging 31 (2010) 4–18. [80] I.T. Ozbolat, B. Koc, Modeling of spatially controlled biomolecules in three-
[45] J.C.K. Wells, M.S. Fewtrell, Measuring body composition, Arch. Dis. Child. 91 dimensional porous alginate structures, J. Med. Device 4 (2010) 41003.
(2006) 612–617. [81] A.N. Ahsan, M.A. Habib, B. Khoda, Resource based process planning for additive
[46] J.J. Ballyns, L.J. Bonassar, Image-guided tissue engineering, J. Cell. Mol. Med. 13 manufacturing, Comput. Des. 69 (2015) 112–125.
(2009) 1428–1436. [82] D. Yoo, Heterogeneous minimal surface porous scaffold design using the distance
[47] E. Terreno, D.D. Castelli, A. Viale, S. Aime, Challenges for molecular magnetic field and radial basis functions, Med. Eng. Phys. 34 (2012) 625–639.
resonance imaging, Chem. Rev. 110 (2010) 3019–3042. [83] W. Mickel, G.E. Schröder-Turk, K. Mecke, Tensorial Minkowski functionals of
[48] J.M. Theysohn, S. Maderwald, O. Kraff, C. Moenninghoff, M.E. Ladd, S.C. Ladd, triply periodic minimal surfaces, Interface Focus 2 (2012) 623–633.
Subjective acceptance of 7 T MRI for human imaging, Magn. Reson. Mater. Phys. [84] S.C. Kapfer, S.T. Hyde, K. Mecke, C.H. Arns, G.E. Schröder-Turk, Minimal surface
Biol. Med. 21 (2007) 63–72. scaffold designs for tissue engineering, Biomaterials 32 (2011) 6875–6882.
[49] X. Wu, L.A. Schepartz, Application of computed tomography in paleoanthropo- http://dx.doi.org/10.1016/j.biomaterials.2011.06.012.
logical research, Prog. Nat. Sci. 19 (2009) 913–921. [85] S. Rajagopalan, R.A. Robb, Schwarz meets Schwann: design and fabrication of
[50] M. Ventura, Y. Sun, V. Rusu, P. Laverman, P. Borm, A. Heerschap, et al., Dual biomorphic and durataxic tissue engineering scaffolds, Med. Image Anal. 10
contrast agent for computed tomography and magnetic resonance hard tissue (2006) 693–712.
imaging, Tissue Eng. Part C Methods 19 (2012) 405–416. [86] L. Elomaa, S. Teixeira, R. Hakala, H. Korhonen, D.W. Grijpma, J.V. Seppälä,
[51] A. Appel, M.A. Anastasio, E.M. Brey, Potential for imaging engineered tissues with Preparation of poly(ε-caprolactone)-based tissue engineering scaffolds by stereo-
X-ray phase contrast, Tissue Eng. Part B Rev. 17 (2011) 321–330. lithography, Acta Biomater. 7 (2011) 3850–3856.
[52] L.A. Feldkamp, L.C. Davis, J.W. Kress, Practical cone-beam algorithm, J. Opt. Soc. [87] F.P.W. Melchels, J. Feijen, D.W. Grijpma, A poly(d,l-lactide) resin for the
Am. A. 1 (1984) 612–619 (=). preparation of tissue engineering scaffolds by stereolithography, Biomaterials 30
[53] H.S. Tuan, D.W. Hutmacher, Application of micro CT and computation modeling (2009) 3801–3809.
in bone tissue engineering, CAD Comput. Aided Des. 37 (2005) 1151–1161. [88] M. Afshar, A.P. Anaraki, H. Montazerian, J. Kadkhodapour, Additive manufac-
[54] J.H. Rothe, I. Rudolph, N. Rohwer, D. Kupitz, B. Gregor-Mamoudou, T. Derlin, turing and mechanical characterization of graded porosity scaffolds designed
et al., Time course of contrast enhancement by micro-CT with dedicated contrast based on triply periodic minimal surface architectures, J. Mech. Behav. Biomed.
agents in normal mice and mice with hepatocellular carcinoma: comparison of Mater. 62 (2016) 481–494.
one iodinated and two nanoparticle-based agents, Acad. Radiol. 22 (2015) [89] S. Khalil, W. Sun, Biopolymer deposition for freeform fabrication of hydrogel
169–178. tissue constructs, Mater. Sci. Eng. C 27 (2007) 469–478.
[55] A.F, D.B.D, H.N. Cardinal, Three-dimensional ultrasound imaging, Phys. Med. [90] J.P. Li, J.R. de Wijn, C.A. van Blitterswijk, K. de Groot, The effect of scaffold
Biol. 46 (2001) R67. architecture on properties of direct 3D fiber deposition of porous Ti6Al4V for
[56] V. Ntziachristos, Going deeper than microscopy: the optical imaging frontier in orthopedic implants, J. Biomed. Mater. Res. Part A 92A (2010) 33–42.
biology, Nat. Methods 7 (2010) 603–614. [91] B. Starly, W. Sun, Internal scaffold architecture designs using lindenmayer
[57] P.N.T. Wells, H.-D. Liang, Medical ultrasound: imaging of soft tissue strain and systems, Comput, Aided Des. Appl. 4 (2007) 395–403.
elasticity, J. R. Soc. Interface 8 (2011) 1521–1549. [92] O.Y, S.-F.L, B. Starly, A Lindenmayer system-based approach for the design of
[58] L.W. Dobrucki, A.J. Sinusas, PET and SPECT in cardiovascular molecular nutrient delivery networks in tissue constructs, Biofabrication 1 (2009) 45004.
imaging, Nat. Rev. Cardiol. 7 (2010) 38–47. [93] R. Gaebel, N. Ma, J. Liu, J. Guan, L. Koch, C. Klopsch, et al., Patterning human
[59] C.-Y. Lin, Y.-H. Chang, C.-Y. Kao, C.-H. Lu, L.-Y. Sung, T.-C. Yen, et al., stem cells and endothelial cells with laser printing for cardiac regeneration,
Augmented healing of critical-size calvarial defects by baculovirus-engineered Biomaterials 32 (2011) 9218–9230.
MSCs that persistently express growth factors, Biomaterials 33 (2012) [94] J.I. Rodríguez-Dévora, B. Zhang, D. Reyna, Z. Shi, T. Xu, High throughput
3682–3692. miniature drug-screening platform using bioprinting technology, Biofabrication 4
[60] J. Kim, Y. Piao, T. Hyeon, Multifunctional nanostructured materials for multi- (2012) 035001.
modal imaging, and simultaneous imaging and therapy, Chem. Soc. Rev. 38 [95] Y. Jin, Y. He, G. Xue, J. Fu, A parallel-based path generation method for fused
(2009) 372–390. http://dx.doi.org/10.1039/B709883A. deposition modeling, Int. J. Adv. Manuf. Technol. 77 (2015) 927–937.
[61] M. McCormick, X. Liu, J. Jomier, C. Marion, L. Ibanez, ITK: enabling reprodu- [96] I.T. Ozbolat, A. Khoda, Design of a new parametric path plan for additive
cible research and open science, Front. Neuroinform. 8 (2014) 13. manufacturing of hollow porous structures with functionally graded materials, J.
[62] T.B. Rowe, Z.-X. Luo, R.A. Ketcham, J.A. Maisano, M.W. Colbert, X-ray computed Comput. Inf. Sci. Eng. 14 (2014) 41005.
tomography datasets for forensic analysis of vertebrate fossils, Sci. Data 3 (2016) [97] D. Qiu, N.A. Langrana, Void eliminating toolpath for extrusion-based multi-
160040. material layered manufacturing, Rapid Prototyp. J. 8 (2002) 38–45.
[63] H. Zhang, J.E. Fritts, S.A. Goldman, Image segmentation evaluation: a survey of [98] K. Samanta, I.T. Ozbolat, B. Koc, Optimized normal and distance matching for
unsupervised methods, Comput. Vis. Image Underst. 110 (2008) 260–280. http:// heterogeneous object modeling, Comput. Ind. Eng. 69 (2014) 1–11.
dx.doi.org/10.1016/j.cviu.2007.08.003. [99] Y. Yu, K.K. Moncal, J. Li, W. Peng, I. Rivero, J.A. Martin, et al., Three-dimensional
[64] D.L. Pham, C. Xu, J.L. Prince, Current methods in medical image segmentation, bioprinting using self-assembling scalable scaffold-free “tissue strands” as a new
Annu. Rev. Biomed. Eng. 2 (2000) 315–337. bioink, Sci. Rep. 6 (2016) 28714.
[65] S. Skoog, P. Goering, R. Narayan, Stereolithography in tissue engineering, J. [100] V. Mironov, Y.D. Khesuani, E.A. Bulanova, E.V. Koudan, V.A. Parfenov,
Mater. Sci. Mater. Med. 25 (2014) 845–856. A.D. Knyazeva, et al., Patterning of tissue spheroids biofabricated from human
[66] M. Whitaker, The history of 3D printing in healthcare, Bull. R. Coll. Surg. Engl. 96 fibroblasts on the surface of electrospun polyurethane matrix using 3D bioprinter,
(2014) 228–229. Int. J. Bioprint. 2 (2016).
[67] D. Mies, W. Marsden, S. Warde, Overview of additive manufacturing informatics: [101] S.M. Giannitelli, D. Accoto, M. Trombetta, A. Rainer, Current trends in the design
“a Digital Thread”, Integr. Mater. Manuf. Innov. 5 (2016) 1–29. of scaffolds for computer-aided tissue engineering, Acta Biomater. 10 (2014)
[68] R. Devillard, E. Pagès, M.M. Correa, V. Kériquel, M. Rémy, J. Kalisky, et al., 580–594.
Chapter 9 – Cell Patterning by Laser-Assisted Bioprinting, in: Methods Cell Biol., [102] A. Ovsianikov, B.N. Chichkov, M.A.K. Liebschner (Ed.)Three-Dimensional
2014, pp. 159–174. Microfabrication by Two-Photon Polymerization Technique, Humana Press,
[69] Y. Lu, S. Chen, A.K.M. Liebschner (Ed.)Computer-Aided Tissue Engineering, Totowa, NJ, 2012, pp. 311–325.

13
I. Ozbolat, H. Gudapati Bioprinting 3 (2016) 1–14

[103] M. Sortino, S. Belfio, B. Motyl, G. Totis, Compensation of geometrical errors of three-dimensional tissue engineering, Biofabrication 7 (2015) 31002.
CAM/CNC machined parts by means of 3D workpiece model adaptation, Comput. [112] P.A. Fleming, W.S. Argraves, C. Gentile, A. Neagu, G. Forgacs, C.J. Drake, Fusion
Des. 48 (2014) 28–38. of uniluminal vascular spheroids: a model for assembly of blood vessels, Dev. Dyn.
[104] C. Xu, W. Chai, Y. Huang, R.R. Markwald, Scaffold-free inkjet printing of three- 239 (2010) 398–406.
dimensional zigzag cellular tubes, Biotechnol. Bioeng. 109 (2012) 3152–3160. [113] C. Kucukgul, S.B. Ozler, I. Inci, E. Karakas, S. Irmak, D. Gozuacik, et al., 3D
[105] C. Xu, K. Christensen, Z. Zhang, Y. Huang, J. Fu, R.R. Markwald, Predictive bioprinting of biomimetic aortic vascular constructs with self-supporting cells,
compensation-enabled horizontal inkjet printing of alginate tubular constructs, Biotechnol. Bioeng. 112 (2015) 811–821.
Manuf. Lett. 1 (2013) 28–32. [114] B.G. Sengers, C.W. Oomens, F.P. Baaijens, An integrated finite-element approach
[106] I.T. Ozbolat, H. Chen, Y. Yu, Development of “multi-arm bioprinter” for hybrid to mechanics, transport and biosynthesis in tissue engineering, J. Biomech. Eng.
biofabrication of tissue engineering constructs, Robot. Comput. Integr. Manuf. 30 126 (2004) 82–91.
(2014) 295–304. [115] S. Bersini, J.S. Jeon, G. Dubini, C. Arrigoni, S. Chung, J.L. Charest, et al., A
[107] T.J. Hinton, Q. Jallerat, R.N. Palchesko, J.H. Park, M.S. Grodzicki, H.-J. Shue, microfluidic 3D in vitro model for specificity of breast cancer metastasis to bone,
et al., Three-dimensional printing of complex biological structures by freeform Biomaterials 35 (2014) 2454–2461.
reversible embedding of suspended hydrogels, Sci. Adv. 1 (2015). [116] P. a. Janmey, J.P. Winer, J.W. Weisel, Fibrin gels and their clinical and
[108] J.E. Trachtenberg, P.M. Mountziaris, J.S. Miller, M. Wettergreen, F.K. Kasper, bioengineering applications, J. R. Soc. Interface 6 (2009) 1–10.
A.G. Mikos, Open-source three-dimensional printing of biodegradable polymer [117] J. Campbell, I. McGuinness, H. Wirz, A. Sharon, A.F. Sauer-Budge, Multimaterial,
scaffolds for tissue engineering, J. Biomed. Mater. Res. Part A 102 (2014) and Multiscale three-dimensional bioprinter, J. Nanotechnol. Eng. Med. 6 (2015)
4326–4335. 21005.
[109] D. Loessner, C. Meinert, E. Kaemmerer, L.C. Martine, K. Yue, P.A. Levett, et al., [118] C. Gomez, A Unit Cell Based Multi-scale Modeling and Design Approach for
Functionalization, preparation and use of cell-laden gelatin methacryloyl-based Tissue Engineered Scaffolds, Drexel University, Philadelphia, PA, 2007.
hydrogels as modular tissue culture platforms, Nat. Protoc. 11 (2016) 727–746. [119] D.H. Turnbull, S. Mori, MRI in mouse developmental biology, NMR Biomed. 20
[110] V. Mironov, R.P. Visconti, V. Kasyanov, G. Forgacs, C.J. Drake, R.R. Markwald, (2007) 265–274.
Organ printing: tissue spheroids as building blocks, Biomaterials 30 (2009) [120] L.A. Hockaday, K.H. Kang, N.W. Colangelo, P.Y.C. Cheung, B. Duan, E. Malone,
2164–2174. et al., Rapid 3D printing of anatomically accurate and mechanically heterogeneous
[111] A. Akkouch, Y. Yu, I.T. Ozbolat, Microfabrication of scaffold-free tissue strands for aortic valve hydrogel scaffolds, Biofabrication 4 (2012) 035005.

14

Anda mungkin juga menyukai