Добавил:
kiopkiopkiop18@yandex.ru t.me/Prokururor I Вовсе не секретарь, но почту проверяю Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_585_Библиотеки_им_академика_М_И_Перельмана.pdf
Скачиваний:
0
Добавлен:
30.08.2026
Размер:
75 Мб
Скачать
4 Introduction to3D Visualization ofAbdominal CT Images
113
the segmentation results, so that 3D reconstruction can be performed after segmentation. However, it should be noted that there are multiple segmentation results in memory after multiple segmentation. Those unnecessary volumetric data that have been segmented should be released in time accord­ing to the actual situation. The interface of region growing and threshold segmentation are shown in Fig.4.14.
IntnDx [] = {0, 1, 0, -1}; // four neighbors
intnDy[]={-1,0,1,0j;
/ / Dene the queue and save coordinates X and Y
Int*pnGrowQueX= new int[nDatax*nDatay];
Int*pnGrowQueY=new int[nDatax*nDatay];//nDatax
and nDatay represent the width and height
respectively
/ / Dene the start and end of the queue. When
Start = End, there is only one point in the queue
Int Start=O, End=0;
/ / Put the seed coordinate onto the front of the
queue
pnGrowQueX[End]=seed_X;
pnGrowQueY[End]=seed_Y;
While(Start<=End)//This while-loop is used for
region growing
{
CurrX=pnGrowQueX[Start];
CurrY = pnGrowQueY[Start];
For(k=0;k<4;k++)//Go through 4 neighbors of the
current point.
{
XX=CurrX+nDx[k];
YY=CurrY+nDy[k];
If ((X) >=0&&xx<nDatax) & & (yy>=0&&yy<nDatay)
/ /Ensure that the data will not cross-border {
CurrentCord=YY*nDatax+xx:
/ / Get the pixel value of the corresponding
position of the source image and the target image
respectively
CurrentResultValue=pResultScal
ars->GetValue(CurrentCord);
CurrentSourceValue=)
pSourceScalars->GetValue(CurrentCord);
/ / The target image is identied as backcolor
in advance. When the area meets the conditions
If((CurrentResultValue==backcolor)&&
(nSeedVal-CurrentSourceValue)<=Threshold))
{
// the pointer at the end of the queue moves
one bit back and adds a point.
End++;
NTotalPixelVal+=CurrentSourceValue;
PnGrowQueX[End]=xx;
PnGrowQueY[End]=yy;
pResultScalars->SetValue(CurrentCord,foreco
lor);//Set the pixel (xx, yy) to forecolor,
indicating that the point is to be merged
nSeedVal=nTotalPixelVal/(End+1);//The gray
average of the previously divided regions
}
}
}
Start++;
} //while
Delete[]pnGrowQueX:
Delete[]pnGrowQueY:
4.3.4 The Module for3D Reconstruction
In this module, contours, MC algorithms, maximum density projection algorithms, and ray casting methods are imple­mented using the functions provided by VTK. Moreover, scaling, rotation, changing the color, adjusting the trans­parency, and segmenting tissue can be implemented on the reconstructed model. The user interface is shown in Fig.4.15.
When contour and marching cube algorithms are used for surface rendering, two functions are mainly involved: VtkContour Filter, the function for contour, and vtkMarch­ingCubes, the function for marching cubes. The key for unsegmented images is to set the value of the iso-surface. When the value is different, the reconstructed tissue is differ­ent. The specic steps by using MC method are as follows:
• Use the vtkBMPReader class in VTK to read BMP les
and save them in the vtkStructuredPoints class.
• Use the vtkMarchingCubes class to set the value of the
contour surface and extract the contour of interest.
• Call vtkPolyDataMapper to map the data processed by
vtkMarchingCubes to geometric data.
• Dene vtkActor, specify information such as lighting,
view, and focal point of the scene, and then use the
vtkRender class to render the entities in the scene.
In the volume rendering algorithm, maximum inten­sity projection algorithm and ray synthesis algorithm are used, mainly involving vtkVolumeRayCastMIPFunction and vtkVolumeRayCastCompositeFunction. The effects achieved by these two algorithms are quite different. For the ray synthesis algorithm, how to set the mapping function is the key which determines the effect, that is, how to set the mapping relationship between grayscale and transparency and color values. VTK mainly involves three classes: vtkVol­umeRayCastCompositeFunction, vkt piecewise Function, and vtkColorTransferFunction. Among them, vtkVolumeR­ayCastCompositeFunction denes a ray synthesis function,
114
S. Bao et al.
Fig. 4.14 Segmentation interface. (a) Interface of regional growth algorithm; (b) interface of threshold segmentation algorithm
a
b
and vtkPiecewiseFunction denes the mapping relationship between CT values and transparency values in a piecewise manner, while vtkColorTransferFunction is a transformation function that denes the CT value and the color value.
The specic steps for using ray casting are as follows:
• Read the sequence BMP les with the vtkBMPReader class in VTK, and store their information in the vtkStruc­turedPoints class.
• Set its CT value to transparency and color value mapping, the vtkPiecewiseFunction class and vtkColorTransfer­Function class are mainly used.
• Use the vtkVolumeRayCastMapper class to achieve data mapping.
• Dene the vtkVolume Actor, specify information such as scene lighting, view, and focus, and then use the vtkRen­der class to render the entities in the scene.
4.3.5 The Module for3D Model Exporting
MI-3DVS integrates image segmentation with 3D recon­struction to form a simple 3D visualization system. The system meets the basic requirements of data visualization of the abdomen, such as 3D visualization of liver, blood vessel,
4 Introduction to3D Visualization ofAbdominal CT Images
115
Fig. 4.15 Reconstruction interface
and spleen. Three-dimensional reconstructed models can be saved in this module, so that these models can be brought up directly afterward without the need for re-segmentation and reconstruction. The saved formats include STL, PLY, and OBJ.At the same time, the module also realizes the function of saving any image data of the volume data. However, for some complex pipelines, such as the gastrointestinal tract, the system is still unable to separate them well.

References

Cirne MVM, Pedrini H.Marching cubes technique for volumetric visu-
alization accelerated with graphics processing units. J Braz Comput Soc. 2013;19:223–33.
Kang SH, Won Y, Lee K, Youn SI, Min S-H, Park YS, Ahn S-H, Kim
H-H.Three-dimensional (3D) visualization provides better outcome than two-dimensional (2D) visualization in single-port laparoscopic distal gastrectomy: a propensity-matched analysis. Langenbeck’s archives of surgery; 2020.
Nicolas G, Fabio C, Daniel A, Reto S, Guoyan Z, Philipp F.Evaluation
of CT-MR image registration methodologies for 3D preoperative planning of forearm surgeries. J Orthop Res. 2020;38(9):1920–30.
Pattana S, Sakuntam S, Kentaro O.Selective-area growth and charac-
terization of cubic GaN grown by metalorganic vapor phase epitaxy. Thin Solid Films. 2020;709:138205.
Santos J, Oliveira MR, Arrais R, Veiga G.Autonomous scene explora-
tion for robotics: a conditional random view-sampling and evalu­ation using a voxel-sorting mechanism for efcient ray casting. Sensors. 2020;20(15):4331.
Wang J, Huang Z, Yang X, Jia W, Zhou T.Three-dimensional recon-
struction of jaw and dentition CBCT images based on improved marching cubes algorithm. Procedia CIRP. 2020;89:214–21.
Application of3D Printing Technology inHepato-Biliary-Pancreatic Surgery
ChihuaFang andZhaoshanFang
5

5.1 Introduction

Although 3D printing was already proposed in the nineteenth century to produce topographic maps layer by layer, the rst real attempts to generate objects that way were made in the 1980s. Swainson of Denmark proposed a process to directly fabricate a product by selective 3D polymerization of a photosensitive polymer (Swainson 1977). In 1979, Nakagawa of Tokyo University reported the use of lamination techniques to produce actual tools (Nakagawa et al. 1979). In 1981, Hideo Kodama of the Nagoya Municipal Industrial Research Institute (Nagoya, Japan) rst proposed the protocol of a functional photopolymer rapid prototyping system (Kodama
1981). The United States and Japan pioneered the practical
development of real 3D printing technology. In 1986, Charles Hull founded 3D Systems, Inc., and this company developed the stereolithography (STL) le format; in 1988, 3D Systems introduced the world’s rst commercial 3D printing system, the SLA-250 (Ventola 2014). It was in the year 1988 that Scott Crump invented and patented a new 3D printing method called Fused Deposition Modeling (FDM); Crump went on to found Stratasys, Inc., and this company developed the rst FDM 3D printer in 1992 (Bagaria etal. 2018).
Traditional manufacturing mainly uses the principles of mechanics, temperature, and pressure. Generally, the technical process of production can be divided into cold scrap removing and thermal deformation processes. Common operations including shearing, grinding, corrosion, and melting are used to remove the redundant parts and to obtain the shape of components. These individual components are then assembled and welded into nal products. Additive manufacturing represented by 3D printing technology and
C. Fang (*) Zhujiang Hospital, Southern Medical University, Guangzhou, China
Z. Fang The Fifth Afliated Hospital of Guangxi Medical University, Nanning, China
rapid prototyping is an “integrated” high-tech developed in the 1980s, which combines computer, CNC technology, laser technology, CAD/CAM, and materials science in one. Compared with traditional manufacturing methods which involve removing parts of a block of material to form the desired shape, 3D printing technology produces far less material waste since it only uses the material necessary to create a part. The manufacturing cycle by conventional methods usually takes a long time, and the shape of sloped surfaces and deep slots are difcult to produce; while addi­tive manufacturing, also known as 3D printing makes up for the shortcomings of the above traditional manufacturing pro­cesses, showing its tremendous advantages of “green technology.”
5.1.1 Principles andConcepts of3D Printing
3D printing, also known as rapid prototyping, refers to advanced technology for manufacturing 3D objects by stack­ing layers of dened sheet materials such as metal and plas­tic. 3D printing technology is mainly used to “manufacture” products by means of layering processing and superposition through computer control, and its core is the combination of digital, intelligent manufacturing, and materials science. 3D printing is fundamentally different from traditional manufac­turing. It has opened a new era of manufacturing and has been hailed as the foundation of the “Third Industrial Revolution” by The Economist.
5.1.1.1 Principles andWorkow of3D Printing
Principles of3D Printing
3D printing involves a rapid prototyping device generating 3D objects by successively “printing” thin layers through stacking sheets of paper and photocuring technology. The mechanism of 3D printing resembles that of the ink-jet printer. A 3D model is rst produced using 3D design soft­ware, and then the drawings are bundled up into a G-code
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 C. Fang, W. Y. Lau (eds.), Biliary Tract Surgery, https://doi.org/10.1007/978-981-33-6769-2_5
117
118
C. Fang and Z. Fang
le, the native language of a 3D printer. The G-code is then sent to the 3D printer where it is generated into a 3D printout. In 3D printing, the printer generates the object by adding layer upon layer of material until the shape of the object is formed. The object can be produced using various adhesive printing materials, including metal or plastics.
3D Printing Workow
There are four typical phases to obtain a physical object by 3D printing technology: modeling, slicing, printing, and post-processing.
Modeling There are two main methods of 3D modeling for 3D printing. One is to directly create 3D digital models using software such as AutoCAD, 3Dmax, and Blender; the other is to rst acquire 3D data of the object using a 3D scanner such as Polhemus, 3D CaMega, and Z Corp., and then those data are processed to generate a digital 3D model.
Slicing
In 3D printing, the virtual 3D model needs to be
sliced into corresponding 2D graphics and then the 2D graphic information is printed. The thickness of the slice is determined by the properties of printing materials and speci­cations of the printer.
Printing There are a continually expanding group of pro­cesses used in 3D printing. Essentially building up layers and fusing, either with an adhesive, or by applying energy to fuse the substrate. The substrate may be in a vat if liquid, in a bed of powder, or extruded from a ‘print head’. Applied energy is in the form of heat, where the substrate is either extruded in molten form by the printer and fused in situ, or by the application of energy such as infra-red, UV light, Laser or Electron beam. Solid objects can also be produced by materials such as alloy powders applied layer by layer and sintered in situ.
Fused Deposition Modeling
By using this technology, a continuous lament of thermo­plastics such as acrylonitrile–butadiene–styrene nylon and wax, is heated in a liqueer into a semiuid state. Under the control of a computer, a wide variety of these semiuid materials are extruded by a 3D printer extruder according to the cross-sectional prole information. After the object is solidied, a single layer of the desired model is formed; by deposing material layer by layer, the nal product is achieved. A representative company using such technology is Stratasys.
Stereolithography
SLA uses photopolymer resin as the printing material. A computer-guided ultraviolet (UV) laser is used to scan the photopolymer resin. The print is gradually lifted and new layers are printed (xed by the UV laser) from below. Since the photopolymer resin is sensitive to UV light, the resin becomes photochemically solidied, stacking layer upon layer of material to form the nal object. Generally, the material used is a liquid photopolymer resin, and the representative company is 3D Systems.
Selected Laser Sintering
Under laser irradiation, the powdered materials become sin­tered. According to the information of interface prole, selective sintering is carried out under the control of computer to accumulate and shape the parts. The typical working materials for this process include metal powder, ceramic powder, and thermoplastics. The representative companies include 3D Systems and EOS.
Direct Metal Laser Sintering
DMLS uses a high-wattage laser to melt and fuse a layer of alloys and metallic powders together, and the process should be repeated layer after layer until the part is completed. This technique is based on alloy metals. Representative enterprises include EOS and MT.
Post-processing After printing, residues of printing materi-
als can cling or remain in the model, and the object normally comes out of the printer with burrs uneven surface. In view of these issues, excess powders can be removed manually, and the burrs and rough surface can be polished; afterward, the 3D physical model should be glued to enhance the hard­ness; nally, coloring should be performed before a nal product is achieved.
5.1.1.2 Types of3D Printing Technologies
Currently, 3D printing mainly includes fused deposition modeling (FDM), stereolithography (SLA), selected laser sintering (SLS), direct metal laser sintering (DMLS), laminated object manufacturing (LOM), electron beam melt­ing (EBM), and three-dimensional printing (3DP).
Laminated Object Manufacturing
In LOM, the working material consists of a roll of thin “lam­inate.” A laser cuts the outline from a slice of the original image. A heated roller presses the laminate cutting onto the previous layer, the laminate roll moves forward and the next layer is cut, heat rolled, and the process continues. Materials used include plastic, paper, or aluminum foil. Representative companies include Helisys and Kinergy.
Electron Beam Melting
The EBM process manufactures parts by melting metal pow­der layer by layer with an electron beam under a vacuum. Generally, the working material is a titanium alloy, and the representative company is Arcam AB.
5 Application of3D Printing Technology inHepato-Biliary-Pancreatic Surgery
119
Three-Dimensional Printing
3DP is a printing process using a droplet ejection method. Through an ink-jet printing head, a liquid binder is sprayed into a layer of powder, and then the binder is solidied, forming a solid layer. This layer-by-layer process repeats until the desired model is formed. Generally speaking, colored plaster is used as the representative material, and the representative company is 3D Systems.
5.1.1.3 Technical Requirements ofMedical 3D
Printing
Only through high-quality original CT images and high­delity 3D reconstructed model can we obtain a high­precision and high-delity printed physical model. Therefore, there are three technical requirements regarding the printing of a 3D physical model for hepatobiliary and pancreatic diseases.
Acquisition ofHigh-Quality Thin-Slice CT Images inDICOM Format
The quality of CT data can be affected by (a) the parameters of original data acquisition, (b) the quality, dose, and injection speed of contrast agent, and (c) module function of different software. Therefore, clinicians, radiologists, and technologists should work together to optimize parameters, so as to collect high-quality thin-slice CT images (Thin slices, as used herein, refers to medical images with a thickness of 0.625–1.5mm), so as to obtain CT data with better contrast (good signal-to-noise ratio), which is very important in 3D modeling. The quality of CT data directly affects the accuracy of the subsequent 3D visualization model of liver, bile, and pancreas.
Construction ofHigh-Fidelity 3D Models
CT value is the basis of images. The 3D modeling process may inevitably lose details in the original raw data, which leads to distortion of the 3D reconstruction accordingly. Using 3D visualization technology, 3D reconstruction is performed by preprocessing, image segmentation, surface rendering, and volume rendering of the original DICOM data. This secondary processing of the original 2D images may lead to some degree of image distortion. In order to reduce image distortion, it is necessary for a Hepato-biliary­pancreatic surgeon who has mastered the 3D reconstruction software and who is competent and experienced in lm reading to analyze the image and reconstruct the 3D model. Only in this way can the 3D printed model be faithful to the original data of CT.
High-Performance 3D Reconstruction Software
The performance of the 3D reconstruction software affects the delity of reconstruction. 3D visualization technology involves multiple steps including image data preprocessing,
segmentation, registration, 3D reconstruction, and visualization display; and it is related to different algorithms or methods of computer image and graphics processing. The processing or calculation of the above different steps may result in the loss of original data, thereby affecting the precision of the reconstruction. Therefore, the requirements of high-delity for 3D printed objects can be satised by optimizing the algorithms of various reconstruction methods and improving the delity of 3D reconstruction.
5.1.1.4 Benets andDrawbacks of3D Printing inMedical Applications
Advantages of3D Printing
High printing accuracy, constant prototyping and enhanced productivity, customization and personalization, and cost reduction.
Disadvantages of3D Printing
Challenges of 3D printing include limited size of objects, limited raw materials, and restriction on printing high­precision instruments.
There are potentially deciencies in medical 3D printing, especially in 3D printing of hepatobiliary and pancreatic models:
• CT image quality affects the quality of the printed model.
3D printed models can be produced by CT-based
volumetric medical images. Loss of details in DICOM
images may inevitably lead to distortion in 3D modeling,
which in turn directly results in information loss in the
preparation stage of printing. The above three situations
may cause different degrees of distortion in the 3D printed
object.
• The 3D printer and printing materials affect the quality of
the model. The performance of the printer affects the
quality of the printed model; the type, texture, and
properties of printing materials also affect the performance,
transparency, tissue elasticity, and stability of 3D printed
objects.
• 3D printing is expensive and time-consuming. Igami etal.
(2014) believed that it takes about 18h and approximately
50,000 ¥ to print a 3D liver at a scale of 70%; the printing
time and cost could increase twofold; moreover, 2–3days
are required for post-processing. Zein etal. (2013) pointed
Printing a liver model can take up to 2days plus one more
for post-processing (Kuroda etal. 2017), and the cost of
printing material ranges between $500 and $800. The cost
in time and money are factors that restrict the conventional
application of 3D printing in hepatobiliary and pancreatic
surgery. 3D printing is only suitable for some screened
patients, not for emergency patients.
120
C. Fang and Z. Fang
5.1.1.5 Applications of3D Printing
In the current scenario, 3D printing has facilitated advances in industrial manufacturing, custom art, aerospace, and biomedical industries. “Urbee,” the world’s rst 3D printed car was created by Ecologic and Stratasys in November 2010; SULSA, the world’s rst 3D printed aircraft was developed by engineers at the University of Southampton in August 2011; in November 2012, the world’s rst 3D printed articial human liver tissues were created by Scottish scientists; in October 2013, a 3D printed artwork nicknamed “the God of ONO” was successfully auctioned; in November 2013, a Texas company by the name of Solid Concepts has manufactured the world’s rst 3D printed metal gun; In China, a large-scale integral titanium alloy key main load­bearing part has been successfully produced by the Beijing University of Aeronautics and Shenyang Aircraft Design Institute, and this project won the rst National Award for Technological Invention in 2012, which made China the rst country in the world to successfully install such components into engineering applications. These high-tech products have promoted the development of 3D printing technology into a new era.
5.1.2 Medical Applications for3D Printing
The rapid development of 3D printing technology has pro­moted its application in the medical eld. Correspondingly, advances in medical 3D printing technology have made tremendous contributions to clinical repair and treatment, medical model manufacturing, tissue organ regeneration, and drug development and testing.
5.1.2.1 Clinical Repair andTreatment
Medical implants produced by 3D printers can better inte­grate into the human body and improve therapeutic efcacy. In March 2014, three bone tumor patients were treated in Xijing Hospital, The Fourth Military Medical University, Xian, China. Their bone defects, in different locations were repaired by implantation of customized 3D printed titanium alloy prosthesis. This was the rst clinical application of 3D printed titanium scapula prosthesis and clavicle prosthesis in the world, and the rst application of pelvic prosthesis in Asia (Fan et al. 2015). In August of the same year, The Department of Orthopedics of the Peking University Third Hospital completed the rst worldwide treatment for atlanto­axial malignant tumor by 3D printing technology, which has opened up a new way to reconstruct cervical vertebral struc­ture after tumor resection.
5.1.2.2 Medical Model Manufacturing
Medical models have been extensively used in the teaching of basic medicine and clinical trials. However, the traditional
method of producing medical models is complicated and time-consuming and they are easily damaged because most of their raw materials are plaster. In recent years, the use of corpses for medical anatomy teaching has become more and more ethically debated and socially controversial. With the progress of 3D printing technology, printed physical models can be used for anatomical teaching. Medical experimental models not only help avoid the above problems, but also realize the personalized manufacturing of special models according to specic needs. After obtaining DICOM data based on MSCT scanning for different parts of the human body, the data can be reconstructed into le formats such as STL, VRML, and PLY by 3D reconstruction software; any human organs can be replicated by a 3D printer. By using 3D printing technology, any anatomical parts, including upper limbs, hands, coronary arteries, and trachea can be printed into physical models, and these models can provide more information than 2D images.
5.1.2.3 Tissue/Organ Regeneration
How to build replacement tissues and organs remains a chal­lenging problem in clinical medicine? Many end-stage (tumor) patients in need of organ replacement have lost their lives because of the persistent organ shortage. As science and technology continue to evolve, 3D printing of human organs is becoming possible and will moderate the disparity between organ supply and organ demand. 3D printing of some viable organ components has already become possible. In 2013, a US military funded research made a major breakthrough on 3D printing for skin and kidney. More recently, German researchers have produced exible articial blood vessels using 3D printing technology; these vessels can fuse with human tissue, not only to avoid rejection, but also to grow muscle-like tissue. These successful cases suggest that it is becoming possible to address the current and future articial organ shortages by 3D printing.
5.1.2.4 Development andTesting ofDrugs
At present, most of the drug testing is carried out on labora­tory animals. It is difcult to get accurate feedback of phar­macological effects on humans. The human liver, kidney, and specic cell tissues printed by 3D technology for the testing of new drugs can not only truly simulate the human body’s response to drugs, but also obtain accurate test results, which can reduce the initial development cost. In 2013, Organovo, Inc. printed a miniature liver with a depth of 0.5mm and a width of 4mm using a 3D printer. Twenty layers of liver cells and vascular wall cells were printed by a bioprinter in a layer by layer manner. The 3D bioprinted mini livers can perform all of the many vital life functions of a real human liver, including metabolism of proteins and cholesterol, enzyme activation, and excretion of drugs. The undeniable benets of 3D bioprinted liver models are gradually being highlighted
5 Application of3D Printing Technology inHepato-Biliary-Pancreatic Surgery
121
in new drug development and drug toxicology testing. Although several issues remain to be addressed, it is antici­pated that 3D printing will continue to evolve and play an essential role in liver tissue engineering.
5.2 3D Printing forHepato-Biliary­Pancreatic Diseases
The basic steps of 3D printing for surgical diseases of the liver, pancreas, and biliary tract are as follows: upper abdominal thin-layer DICOM data acquisition, 3D digital preparation, 3D physical model printing, and 3D printing post processing (Fig.5.1).
5.2.1 Acquisition ofCT Data
An enhanced thin-layer scanning of the liver, bile, pancreas, spleen, and abdominal vessels in the upper abdomen was per­formed by Multidetector Computed Tomography (MDCT) or an enhanced thin-layer of MRI. Multiphase images (plain scan phase, arterial phase, portal venous phase, and delayed phase) were obtained with an image slice thickness of 0.625–
1.5mm. Data was archived in the DICOM format.

5.2.2 Digital Preparation

Thin-layer CT data were rst processed by a post-processing workstation, and then imported into a 3D visualization software system (such as MI-3DVS, China; Mevis, Germany; or MIMICS, Materialise, Belgium) for program segmentation and reconstruction. Note: careful and accurate reading and analysis of CT images before 3D reconstruction are important to ensure the accuracy of 3D modelling. The DICOM data were analyzed, fused, calculated, segmented, and rendered by a 3D software system. The shape and spatial distribution of the liver, biliary tract, blood vessels, and lesion were described and interpreted; 3D image models of intrahepatic vessels and lesion were obtained and exported to a mesh- type le (stereolithography (STL) le), or in other formats, such as VRML and PLY.These les were prepared for 3D printing. Files in STL format were acquired and further processed by Materialise Magics software in order to (a) assess the accuracy of intersecting vessels and bile ducts arising from subtle time or position artifacts between CT imaging phases, thereby generating nonoverlapping geometric gures, (b) perform hollowed vascular and bile duct structures, and (c) divide liver mesh structure (stereolithography (STL) le) into graft and remnant parts according to the proposed surgical resection plane.
3D data DICOM
3D physical model
Fig. 5.1 Flow chart of 3D printed physical model
3D reconstruction
3D printing post-processing
3D printing
122
ab
C. Fang and Z. Fang
5.2.3 3D Printing forHepato-Biliary­Pancreatic Diseases
The STL les produced through digital preparation were imported into a 3D printer (such as Connex 350 3D printer, Stratasys; or 3D Printer, AGILISTA-3100, Keyence Co.; Spectrum ZTM 510 3D Printer, Z Corporation) for printing 3D physical models of intrahepatic and extrahepatic biliary tract, blood vessels, and lesions (Figs.5.2, 5.3, 5.4, and 5.5). Liver parenchyma can be made of transparent materials such as Tango or Vero, or jelly wax; hepatic vein structure can be Tangoblack or Veroblue; other vessels can be mixed with transparent materials such as Tango or Vero; ZP 150 powder can also be used for printing.
5.2.4 Post-Processing of3D Printing
Post-processing of 3D printed physical models for surgical diseases of the liver, pancreas, and biliary tract mainly involves reprocessing of the support materials and vessels. For example, after 3D physical models are printed by the Connex 350, removing support or excess material from the 3D print is a necessary step; the vessels are injected with color dye by a water gun, and the liver surface is coated to obtain visualized intrahepatic bile duct and vascular structure. After 3D physical models of intrahepatic bile ducts, blood vessels, and lesions are printed by the Spectrum ZTM 510, excess powder of the 3D print should be manually removed; and then the surface of the model should be osmotically
cured using a curing agent Z-Bond90 (3D Systems, USA). The physical model of the hepatic parenchyma shell and the model of the hepatic ducts are assembled to form the casting mold. The liquid transparent wax is injected into the mold. After solidication, the outer casting should be removed to obtain the transparent 3D physical model.
Fig. 5.3 Front view of 3D printed model of hilar cholangiocarcinoma. Note: Green for intrahepatic dilated bile duct system, orange for hilar tumor, dark blue for portal vein system, and red for celiac artery system
Fig. 5.2 3D printed model of intrahepatic vessels and tumors of central HCC. (a) Front view; (b) back view. Note: Dark blue for hepatic vein system, pink for portal vein system, and red for celiac artery system
ab
5 Application of3D Printing Technology inHepato-Biliary-Pancreatic Surgery
123
Fig. 5.4 Front view of 3D printed model of hepatolithiasis. (a) 3D printed model of left intrahepatic calculi; (b) 3D printed models of left and right hepatic calculi. Note: Red for the celiac artery system, dark
blue for the hepatic vein system, green for the dilated bile duct, light blue for the portal venous system, and white for hepatolithiasis
ab
Fig. 5.5 3D printed physical model of ampullary tumor. (a) Front view; (b) back view. Note: Dark brown for the ampullary tumor, dark blue for the portal vein system, green for the dilated bile duct system, and red for the celiac artery system
5.3 Application of3D Printing forComplicated Hepato-Biliary­Pancreatic Surgery
With the development of computer technology, 3D visualiza­tion technology for 3D modelling has become an important auxiliary tool for planning surgically complicated hepatobi­liary and pancreatic surgery. This technology has more
advantages than conventional CT and MRI 2D imaging. However, 3D images reconstructed based on the patient’s MDCT data are usually displayed on a 2D computer moni­tor, so the true depth perception response is limited, and dif­ferent physicians have different perceptions of the spatial anatomical relationship between blood vessels and liver tumors. 3D physical printing of images based on 3D visual­ization has broken through this bottleneck. By observing 3D