- Email: [email protected]
A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels
G Model JINJ 7388 No. of Pages 8
Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels Vaibhav Bagaria* , Kshitij Chaudhary Sir HN Reliance Foundation Hospital, Mumbai, India
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 24 August 2017
Introduction: Preoperative planning is an important aspect of any orthopedic surgery. Traditionally, surgeons mentally rehearse the operation and anticipate problems based on data available from “radiography” like MRI and CT. 3D printed bio-models and tools, or “3Dgraphy” can simplify this mental exercise and provide a realistic and user-friendly portrayal of this radiographic data. Methods: Five surgeons participated in this multicenter study. 3D printed biomodels were obtained for 50 surgical cases that included periarticular trauma (24), pelvic trauma (11), complex primary (7), and revision arthroplasty (8). CT scan data was used to generate computer models which were then 3D printed in real size. These models were used to understand pathoanatomy and conduct simulated surgery as a part of preoperative planning. The models were sterilized and were used for intraoperative referencing. Following each case, the operating surgeon was asked to fill out a structured questionnaire to report on the perceived benefits of these tools. Results: All surgeons reported that the biomodels provided additional information to conventional imaging that enhanced their knowledge of the complex pathoanatomy. It was useful in preoperative planning, rehearsing the operation, surgical simulation, intraoperative referencing, surgical navigation, preoperative implant selection, and inventory management. This probably reduced surgical time and improved accuracy of the surgery. All surgeons reported that they would not only use it themselves but also recommend it to other surgeons. Conclusion: 3Dgraphy was found to be a valuable tool in orthopedic surgeries that involve complex pathoanatomy like pelvic trauma, revision arthroplasty, and periarticular fracture. As the technology evolves and improves, they are likely to become a standard component of many orthopedic procedures. © 2017 Elsevier Ltd. All rights reserved.
Keywords: 3D printing Acetabular fractures Complex fractures Preoperative planning 3dGraphy
Introduction Complex fractures and difficult arthroplasty surgeries require meticulous preoperative planning and careful execution. Preoperative evaluation, appropriate inventory planning, intraoperative availability of desired instrumentation and tools, teamwork, and surgeon's experience all play a role to ensure optimal results. The starting point for this entire preoperative exercise is getting the necessary imaging (radiography) done and plan the execution based on them. Two dimensional “axial” data requires the radiologist and the surgeon to develop a mental image of the pathoanatomy. 3D printed biomodels, or 3Dgraphy, can simplify this mental exercise and provide a realistic and user-friendly portrayal of this radiographic data (Fig. 1).
* Corresponding author at: Department of Orthopedics and Joint Replacement Sir HN Reliance Foundation Hospital Prathna Samaj, Mumbai, India. E-mail address: [email protected] (V. Bagaria).
3D printing has revolutionized many areas of, manufacturing and is increasingly playing an important role in health care [1]. The process of 3D printing involves “layered manufacturing” wherein thin layers of material deposited, which when stacked up create a 3D object. The availability of imaging modalities such as CT scans and MRI have ensured that the cross sectional or axial data of each imaged part is available to us in a format that allows a seamless integration with 3D print technology [2]. Based on these images, 3D printed model of any body part can be created. In cases of complex arthroplasty, production of such models is of immense help to surgeons [3]. It goes beyond the usual 3D reconstructed images form MRI and CT and gives the surgeon an opportunity to study the problem by not just seeing it in two dimension but holding the same in his hand ensuring a real-size 3D perspective. Of various advantages, ability to assess the bone defects, fracture patterns if any, position of previous implants in revision cases and plate contouring are the most apparent. The biomodel gives an opportunity to plan for the required instrumentation and implants thus optimizing the inventory. The surgeons can
http://dx.doi.org/10.1016/j.injury.2017.08.058 0020-1383/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
2
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
The clinical and surgical details were retrospectively reviewed. The generation of the 3D biomodel involved the following stages: (1) Image acquisition and denoising, (2) Segmentation, (3) Generation of a 3D CAD model, (4) Postprocessing of CAD model, and (5) 3D printing of biomodel. Image acquisition and denoising
Fig. 1. Radiography to 3D graphy – the improvisation pyramid for surgical planning and simulation.
simulate the procedure and if required build customized templates and jigs based on the patients disease, anatomy and his preferences [4,5]. While there are isolated case reports, case series, technical notes and review articles on the subject, there are no systemic studies to analyze the impact of these models on surgeons perception and ultimate surgical execution using a structured reporting template. We used 3D printed models (3D POST1) to plan and simulate the procedure in 50 cases of complex periarticular trauma, complex primary and revision arthroplasty. As a part of the study, benefits and the utility of 3D POST was assessed by using a structured surgeon reported templates. Aims 1. Prospective evaluation of surgical cases performed using 3D printed biomodels. 2. To develop guidelines for choosing appropriate “open source” computing algorithms, rapid prototyping technique and raw materials for creating 3D printed biomodels.
Method Study group The study comprised of a consecutive series of 50 patients, for whom the operating surgeon used a 3D printed biomodel for surgical planning and execution. The study was inclusive of all types of orthopaedic surgeries. (Table 1).
Table 1 Number of cases in the series. Region
No. of cases
Acetabulum Proximal humerus Elbow (periarticular) Neck femur Tibial condyle Foot and ankle Periarticular miscellaneous Complex primary hip arthroplasty Revision hip arthroplasty Peri-prosthetic fractures Total
11 4 4 3 6 4 3 7 6 2 50
CT scan was used as the imaging modality in all cases as it best captures bone data. The CT scan was done using the MRCP (Medical Rapid-prototyping Computed-tomography Protocol) guidelines (Table 2) as described earlier by our group, in 38 (76%) cases [6]. The protocol could not be standardized for the rest, as the CT scan was already performed before the surgeon requested for a biomodel to be made. In brief, the MRCP protocol lays down guidelines for CT image acquisition such that is maximizes the resolution of images without significantly increasing the radiation exposure to the patient. CT images were obtained in DICOM format and were transferred either through a cloud server or sent physically on a CD to the model production location. A copy of raw data was always retained as permanent archives for all patients. Denoising: If the images had random noise (especially if previous implant was present), denoising was carried out to prevent metal artefacts in the CAD (computer aided design) models. Thus the first step was sequential denoising of the CT acquired data. We employed the spatial smoothing method, which is a computer algorithm to reduce the noise without loosing anatomical information [7]. Segmentation Segmentation is a process in which the desired area of anatomy is labeled and thus separated from the surrounding undesired structures on axial, sagittal and coronal images. Computer software was used to assign a label to each voxels that was thus selected. We used two methods (software) for generating a model: Osirix [8]: This is a free open source software that has a 3D rendering function called the surface rendering tool. This software has quite limited capability and uses thresholding for segmentation. However, ability to manually modify the segmentation slice by slice does not exist in this software. This was used for the initial 15 cases. As we gained more knowledge about segmentation we switched over to another open-source software developed by Harvard University, namely 3D Slicer. 3D slicer [9]: This software has the capability to use both thresholding and manual segmentation for selection of the desired voxels. For the remaining 35 cases we used this technique to generate the biomodel (Fig. 2). The initial step was to use thresholding to separate high contrast areas such as bone from the surrounding soft tissue. We found that in osteoporotic bones or cancellous bone, a lot of low density bone was not selected. In addition, there were errors due to partial volume effect around intra-articular areas [10]. Therefore, the next step was to manually navigate slice by slice to “paint” the desired areas or “unpaint” unwanted areas of bone. Generation of 3D CAD model After segmentation the model was exported as an STL (Surface Tessellation Language) file. This was subjected to further processing to improve the quality and remove unwanted parts of the 3D model.
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
3
Table 2 MRCP protocol for CT scan image acquisition ideal for creating 3D printed bio models. Parameters
Description
Field of view (FOV) Scout Region of interest (ROI) kV mA Pitch Collimation Slice thickness Slice increment Kernel/algorithm
This is the region of interest. FOV measuring 12 12 inches is adequate Depends on region of interest and helps planning ROI should be identified Automatic Usually automatic 512 512 1.25–1.50 mm 1–1.5 mm 0.625–0.75 (less than 1 mm) Moderate/soft tissue (not to use “bone/detail”)
Fig. 2. Processing of CT acquired date prior to actual printing of 3D Models.
Postprocessing This was done using Meshlab (Visual Computing Lab, Pisa, Italy), Blender (Blender Foundation, Amsterdam, The Netherlands) and Meshmixer (Autodesk Inc, USA). The postprocessing was carried out in the following sequential steps: 1) Down sampling and reducing complexity: The CAD model surface is represented my triangular areas (tessellations) without gaps or overlap. This virtual mesh of triangles can get quite complicated for a complex area, especially with high resolution CT images. This sometimes results in a very large file. Hence, for such files the complexity was reduced to a
manageable level without compromising on the surfaced details of the model. 2) Cleaning: Unwanted segmentation artefacts were removed. In addition, areas of bone that were clinically unimportant were removed to reduce the cost of printing. If any anatomical areas were disjointed, then an artificial bridge was created to connect it to the main structure to preserve the anatomical relationship. Finally, mesh cleaning and healing algorithms were applied to correct any errors. 3) Smoothening: Finally, smoothening algorithm was applied taking care to not lose the anatomic fidelity of the surface (Fig. 3).
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
4
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 3. Postprocessing of STL file (Fig B after smoothening effect).
3D printing of the biomodel
Post surgery assessment
All the models were produced by Timeless Innovation Labs, Mumbai, India using the Wanhao Duplicator Printer with a resolution of 0.1 mm (100 mm). The 3D printing technology used was Fused Deposition Modelling (FDM). FDM produces 3D biomodels by extruding thin strings of molten polymer to form layers as the extruded material hardens instantly post extrusion from the nozzle (Fig. 4) [11]. The polymer used in the study was a thermoplastic resin called ABS (Acrylonitrile Butadiene Styrene). The average time for printing a 6 inch model was 12 h. Once the biomodel was printed, the supports were manually removed and the surface smoothened with a rasp. The models were delivered to the surgeon prior to the surgery. The model was sterilised using Ethylene oxide sterilization
All five surgeons were asked to fill out a standardized questionnaire after the surgery (Fig. 5) where the surgeon rated individual domains within the questionnaire on a scale of 1 to 10 with respect to usefulness of the model. Each biomodel was assessed on 10 criteria and each criterion was given a score between 1 and 10 with 1 being lowest and 10 being highest. Results In all the cases the surgeon found that the model accurately represented the anatomy, helped in preoperative planning and decreased surgical time (Mean score above 8) (Table 2). They also suggested that they would use the model and recommend it to surgeons in future (mean score 8.9). The greatest unhappiness was the time taken for production of model (mean score 4.6). Most surgeons believed that although there may have been some alterations in the way they tackled the cases on based of new information, they did not think that it altered their surgical plan very significantly (mean score 5.7). Rest of the parameters that pertained to inventory management, avoidance of intraoperative complications and postoperative assessment faired above average with mean scores of between 6 and 8 on each parameter (Fig. 6). Following the experience of these cases a streamlined protocol as shown in the flowchart (Fig. 7) was developed for 3D printing biomodels. In contrast to the expensive proprietary software, this protocol was based on freely available open source programs and includes even the fine nuances on how to obtain the appropriate medical imaging for purpose of 3D printing (Table 3). Case examples
Fig. 4. 3D printing of acetabulum being carried out on a Fused deposition modeling (FDM) machine using ABS polymer.
Illustrative case 1: ST a 21 year old case of Juvenile Rheumatoid arthritis had bilateral severe arthritis of the hip with dysplastic changes. Weighing only 21 kg, and being wheelchair bound since age of 6, performing hip arthroplasty was a challenging affair. His
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
5
Fig. 5. Surgeon reported self assessment questionnaire for 3D models.
pelvis was very small and bone very osteoporotic and to compound the problem the anatomy was fairly distorted in the region. (Figure) A CT scan gave an insight on what surgeons could expect during surgery. 3D printed biomodel from this CT Scan post processing (Fig. 8) however allowed the surgeon to simulate the procedure, understand the requirement of a small size cup and order an inventory (Fig. 9). The models proved an invaluable tool in accomplishing a successful surgery. Illustrative case 2: A 25-year-old car accident victim who suffered from a T Type acetabular fracture was scheduled to undergo a surgery. A planning on OSirix was followed by printing the 3D bio model of both injured and mirror image of non injured side. This model gave an accurate visual impression of the fracture pattern (Fig. 10) and also allowed the template of the implant to be contoured according to patients anatomy and fracture pattern using the mirror image model of opposite side (Fig. 11). The surgeon felt that this significantly reduced the surgical time and help accomplish the surgery in desired fashion.
Discussion 3D printing – also known as rapid prototyping, stereolithography, or addititive manufacturing – is ushering in the next technological revolution. These days it is hard not to notice media headlines proclaiming triumphs of surgeons solving complex surgical problems with the aid of 3D printing. Typically newspaper articles will claim that an intrepid surgeon successfully applied recent breakthroughs in 3D printing technology to help his patient. Despite these claims of novelty, it is interesting to know that 3D printing or rapid prototyping technology has been around since the early 1980s [12]. Even a perfunctory glance at the scientific literature in the last decade will reveal how permeative this technology has become, not only in orthopaedic surgery but also in other surgical specialities [13]. In this paper we reveal our journey of developing the applications of this technology for orthopaedic surgery in the last decade (Fig. 12). Traditionally, before the advent of PACS (Picture Archiving and Communication System), the images obtained using advance imaging such as CT and MRI were printed on films losing valuable information in the process. PACS has changed the way we view images adding another dimension to interpretation. With
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
6
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 6. Graphical representation of average score post assessment using the standard questionnaire.
provided by radiologist to generate 3D images as per their specific requirements, allowing them to virtually manipulate these images on their computer screens [14]. Many surgeons feel that the 3D images provided either by the radiology department or reconstructed by themselves using softwares such as Osirix provide almost a true 3D representation of anatomy. However, these 3D computer images are limited by the two-dimensional nature of the computer screen. Depth perception and relative size cannot be comprehended on the computer screen. Real-size 3D printed biomodels are invaluable to study complex anatomy because they can provide that extra dimension. Physical interaction with anatomy, allowing the representative bone to be touched and examined from all angles (tactile stereotaxy) is the true benefit of 3D printed biomodels [13]. In this case series, we have used 3D printed biomodels for complex orthopaedic cases, such as difficult primary and revision arthroplasty, periatricular fractures, and pelvic trauma. The standard of care for these cases has been to study axial images or 2D representation of 3D computer images provided by the radiology department. 3D printed biomodelling for these cases provided the surgeons an opportunity to physically interact (tactile stereotaxy) with the biomodels during the process of preoperative planning. Acetabular fracture surgeries are known to have a steep learning curve. This is even documented in a report by Matta and Merrit where in they showed that for every group of 20 patients operated the experience improved the ability to avoid unsatisfactory reductions and to perform anatomic reductions [15]. The same was echoed by Kebaish, Roy, and Rennie who Table 3 Average scores.
Fig. 7. Flowchart for developing 3D printed models using MRCP protocol and open source free softwares.
expansing of computing prowess, these DICOM images can be easily reconstructed into 3D computer images that allow a more intuitive understanding of pathoanatomy. Technologically savvy orthopaedic surgeons were already manipulating the DICOM data
Parameters
Average score
Accurate representation Turn around time Helpful in preoperative planning Altered the surgical plan Reduced surgical time Improved inventory management Avoiding unforeseen complications Postoperative Xray assessment New information compared to preoperative CT Would recommend its use to other surgeons
8.1 4.6 8.3 5.7 7.8 8.2 6.8 7.6 6.2 8.9
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 8. 3D printed model of pelvis of 21 year old person suffering from Juvenile Rheumatoid arthritis suffering from dysplasia and arthritis of both hips. Given the small size of the pelvis and distorted anatomy, the 3D biomodel proved to be an invaluable tool in planning the surgery.
demonstrated that reductions obtained by less experienced surgeons had significantly lower rates of anatomic reductions [16]. While getting a hands on experience doing live surgeries may be the ideal way of improving ones surgical skills, ability to preoperatively simulate the same by carrying out various surgical steps on an accurate representing 3D printed model of the case is a valuable tool as shown by this study. The study group that comprised of senior level surgeons felt value in performing these simulations and concurred that if given a chance they would recommend to fellow surgeons. In periarticular fracture accurate reduction and implant placement are two key elements. To ensure these, understanding of the anatomy, good intraoperative imaging and optimal spending of the operative time is required. While understanding the fracture pattern is subjective and depends on knowledge and experience, intraoperative fluoroscopy is a two dimensional and entails cumulative radiation exposure risks. It is reported that each minute of exposure (60 shots) is equivalent to one computed radiography exposure or 4rads of radiation and that many peri articular fracture fixation scenarios require many minutes of these exposures [17]. The 3D printed models help understand the
Fig. 9. Simulation being carried out on the 3D model. In this case acetabular reaming was performed and trialing for the cup and stem was done. As was expected, even post reaming the smallest available cemented cup could be implanted.
7
Fig. 10. 3D printed biomodel of a t Type fracture of the acetabulum. Surgeon can visualize the fracture anatomy and decide and design a plan for surgical execution.
complex spatial anatomy of these fractures, help in planning geometric parameters of target plate fixation and avoid wide surgical exposures. The participating surgeons concurred that with the use of these 3D models, the surgical time can be reduced; the amount of bleeding can be decreased, the radiation exposure to surgeon and the operating room team can be reduced and as corollary the chances of wound infection also goes down. While primary total hip replacement has now become a standard procedure well controlled by orthopedic surgeons, there may remain a few primary cases that remain challenging and complex. While dealing with cases such as dysplasia of the hip, arthrodesis and protrusio hip the surgeon have to biomechanics, bone defect and planned reconstructive technique. Similarly in revision scenario, it is important to know the possible cause of failure, strategies implant removal, choose implants and formulate a strategy for adequate reconstruction [18,19]. In both these conditions, having a real sized model of the area helps understand the biomechanics and bone defect better [20]. The surgeons may also be able to carry out simulation by acetabular reaming or femoral rasping to determine the size of the socket and canal. These simulations also provide an insight in the nature of bone quality and strength in a given area. In these studies, we had 10 cases in which the biomodels was used for arthroplasty planning. They proved invaluable and in virtually all cases some alteration in surgical plan was effected because of these 3D printed models. With newer metal printers with capability of printing biocompatible materials now being easily accessible, many companies are
Fig. 11. Printing the mirror image of the opposite hip allowed the surgeon to template and contour the plate to normal anatomy and help restore the injured side as close to its anatomical position as possible.
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
8
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
costs reduce [1]. One also has to consider the time and effort spent by the human capital (radiologists, engineers, or the surgeon’s themselves) in producing these biomodels. However, the 3D printer cost has dramatically reduced over the years and a number of open source softwares are available. This has made the technology accessible in poor resource areas such as our country. The ultimate trial would be to determine the value of biomodels which would be determined by considering both the benefits and costs (value = benefit/cost). Conclusion In all the 50 cases the use of 3D printed biomodel was perceived to have reduced the surgical time, improve the understanding of the condition, reduce the inventory, decrease the surgical time and seen as possibly having a favorable impact on patient outcomes. We propose a protocol for using open source software to generation CAD models which then can be printed with the material of choice. References
Fig. 12. A Post operative X rays showing a well reduced fracture and a contoured plate based on printed model.
now 3D printing the wedges and cones and patient specific hip and knee implants may become much more common in coming years [21]. The ultimate benefit of application of a new technology would be measured in terms of patient outcomes. Preoperative planning using 3D biomodels may help reduce surgical time and estimated blood loss. It also has the potential to reduce complications and improve patient outcomes. It is not hard to imagine how it can do so, especially in surgeries for complex pathology. Although our study is an attempt to generate such evidence, a controlled prospective trial would be required to establish the veracity of these claims. Establishing benefit of 3D biomodeling would require consideration of both the case complexity and surgeon experience and this might prove daunting. Models of simple fractures or pathology may have limited benefit. But as complexity of the fracture increases the information gained from the model may exceed that obtained from radiographs or routine CT scans [1]. The benefit provided by 3D modeling is also dependent on the experience of the surgeon. Less experienced surgeons and trainees would find 3D models of moderately complex fractures useful. As surgeon expertise increases the benefit gained from such models would decrease. However, in extremely complex and unusual fracture patterns, even the most experienced surgeons would derive significant benefit from 3D biomodels [1]. The added expense is an important consideration to justify the use of 3D biomodels. Although we did not do a controlled cost analysis in our series, we can predict some assumptions. The initial cost of setting up a 3D printing unit (printers and buying proprietary software) is high, but with time and experience the
[1] Tetsworth KD, Mettyas T. Overview of emerging technology in orthopedic surgery: what is the value in 3D modeling and printing? Tech Orthop 2016. [2] Esses SJ, Berman P, Bloom AI, Sosna J. Clinical applications of physical 3D models derived from mDCT data and created by rapid prototyping. AJR Am J Roentgenol 2011196(6). [3] Sariali E, Boukhelifa N, Catonne Y, Pascal Moussellard H. Comparison of threedimensional planning-assisted and conventional acetabular cup positioning in total hip arthroplasty: a randomized controlled trial. J Bone Joint Surg 2016;98 (2)108–16, doi:http://dx.doi.org/10.2106/JBJS.N.00753 American Volume. [4] Krishnan SP, Dawood A, Richards R, Henckel J, Hart AJ. A review of rapid prototyped surgical guides for patient-specific total knee replacement. J Bone Joint Surg 2012;94(11)1457–61 British volume. [5] Sariali E, Mauprivez R, Khiami F, Pascal-Mousselard H, Catonné Y. Accuracy of the preoperative planning for cementless total hip arthroplasty: a randomised comparison between three-dimensional computerised planning and conventional templating. Orthop Traumatol 2012;98(2):151–8. [6] Bagaria V, Shah S, Bhansali R. Using 3D-printed patient-optimized surgical tools (3D POST) for complex hip and knee arthroplasty N2. In: Bagaria V, editor. Arthroplasty – A Comprehensive Review. Rijeka: InTech; 2016 Ch. 09. [7] Thilagamani S, Shanthi N. Gaussian and gabor filter approach for object segmentation. J Comput Inf Sci Eng 2014;14(2):021006. [8] Rosset A, Spadola L, OsiriX Ratib O. An open-source software for navigating in multidimensional dICOM images. J Digit Imaging 2004;17(3):205–16. [9] Fedorov A, Beichel R, Kalpathy-Cramer J, et al. 3D slicer as an image computing platform for the quantitative imaging network. Magn Reson Imaging 2012;30 (9):1323–41. [10] Green N, Glatt V, Tetsworth K, Wilson LJ, Grant CA. A practical guide to image processing in the creation of 3D models for orthopedics. Tech Orthop 2016;31 (3):153–63. [11] Gür Y. Additive manufacturing of anatomical models from computed tomography scan data. Mol Cell Biomech MCB 2014;11(4):249–58. [12] Hull C. Apparatus for production of three-dimensional objects by . . . 1996, http://www.google.com/patents/US5556590. [13] Martelli N, Serrano C, van den Brink H, et al. Advantages and disadvantages of 3-dimensional printing in surgery: a systematic review. Surgery 2016;159 (6):1485–500. [14] Reddix RN, Webb LX. Computer-assisted preoperative planning in the surgical treatment of acetabular fractures. J Surg Orthop Adv 2007;16(3):138–43. [15] Matta JM, Merritt PO. Displaced acetabular fractures. Clin Orthop Relat Res 1988;230:83–97. [16] Kebaish AS, Roy A, Rennie W. Displaced acetabular fractures: long-term follow-up. J Trauma 1991;31(11):1539–42. [17] Lo NN, Goh PS, Khong KS. Radiation dosage from use of the image intensifier in orthopaedic surgery. Singapore Med J 1996;37(1):69–71. [18] Robertson DD, Sutherland CJ, Lopes T, Yuan J. Preoperative description of severe acetabular defects caused by failed total hip replacement. J Comput Assist Tomogr 1998;22(3):444–9. [19] Munjal S, Leopold SS, Kornreich D, Shott S. Finn HA: CT-generated 3dimensional models for complex acetabular reconstruction. J Arthroplasty 2000;15(5):644–53. [20] Yu AW, Duncan JM, Daurka JS, Lewis A, Cobb J. A feasibility study into the use of three-dimensional printer modelling in acetabular fracture surgery. Adv Orthop 20152015(4) 617046-4. [21] Wong KC, Kumta SM, Geel NV, Demol J. One-step reconstruction with a 3Dprinted, biomechanically evaluated custom implant after complex pelvic tumor resection. Comput Aided Surg 2015;20(1):14–23.
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels Vaibhav Bagaria* , Kshitij Chaudhary Sir HN Reliance Foundation Hospital, Mumbai, India
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 24 August 2017
Introduction: Preoperative planning is an important aspect of any orthopedic surgery. Traditionally, surgeons mentally rehearse the operation and anticipate problems based on data available from “radiography” like MRI and CT. 3D printed bio-models and tools, or “3Dgraphy” can simplify this mental exercise and provide a realistic and user-friendly portrayal of this radiographic data. Methods: Five surgeons participated in this multicenter study. 3D printed biomodels were obtained for 50 surgical cases that included periarticular trauma (24), pelvic trauma (11), complex primary (7), and revision arthroplasty (8). CT scan data was used to generate computer models which were then 3D printed in real size. These models were used to understand pathoanatomy and conduct simulated surgery as a part of preoperative planning. The models were sterilized and were used for intraoperative referencing. Following each case, the operating surgeon was asked to fill out a structured questionnaire to report on the perceived benefits of these tools. Results: All surgeons reported that the biomodels provided additional information to conventional imaging that enhanced their knowledge of the complex pathoanatomy. It was useful in preoperative planning, rehearsing the operation, surgical simulation, intraoperative referencing, surgical navigation, preoperative implant selection, and inventory management. This probably reduced surgical time and improved accuracy of the surgery. All surgeons reported that they would not only use it themselves but also recommend it to other surgeons. Conclusion: 3Dgraphy was found to be a valuable tool in orthopedic surgeries that involve complex pathoanatomy like pelvic trauma, revision arthroplasty, and periarticular fracture. As the technology evolves and improves, they are likely to become a standard component of many orthopedic procedures. © 2017 Elsevier Ltd. All rights reserved.
Keywords: 3D printing Acetabular fractures Complex fractures Preoperative planning 3dGraphy
Introduction Complex fractures and difficult arthroplasty surgeries require meticulous preoperative planning and careful execution. Preoperative evaluation, appropriate inventory planning, intraoperative availability of desired instrumentation and tools, teamwork, and surgeon's experience all play a role to ensure optimal results. The starting point for this entire preoperative exercise is getting the necessary imaging (radiography) done and plan the execution based on them. Two dimensional “axial” data requires the radiologist and the surgeon to develop a mental image of the pathoanatomy. 3D printed biomodels, or 3Dgraphy, can simplify this mental exercise and provide a realistic and user-friendly portrayal of this radiographic data (Fig. 1).
* Corresponding author at: Department of Orthopedics and Joint Replacement Sir HN Reliance Foundation Hospital Prathna Samaj, Mumbai, India. E-mail address: [email protected] (V. Bagaria).
3D printing has revolutionized many areas of, manufacturing and is increasingly playing an important role in health care [1]. The process of 3D printing involves “layered manufacturing” wherein thin layers of material deposited, which when stacked up create a 3D object. The availability of imaging modalities such as CT scans and MRI have ensured that the cross sectional or axial data of each imaged part is available to us in a format that allows a seamless integration with 3D print technology [2]. Based on these images, 3D printed model of any body part can be created. In cases of complex arthroplasty, production of such models is of immense help to surgeons [3]. It goes beyond the usual 3D reconstructed images form MRI and CT and gives the surgeon an opportunity to study the problem by not just seeing it in two dimension but holding the same in his hand ensuring a real-size 3D perspective. Of various advantages, ability to assess the bone defects, fracture patterns if any, position of previous implants in revision cases and plate contouring are the most apparent. The biomodel gives an opportunity to plan for the required instrumentation and implants thus optimizing the inventory. The surgeons can
http://dx.doi.org/10.1016/j.injury.2017.08.058 0020-1383/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
2
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
The clinical and surgical details were retrospectively reviewed. The generation of the 3D biomodel involved the following stages: (1) Image acquisition and denoising, (2) Segmentation, (3) Generation of a 3D CAD model, (4) Postprocessing of CAD model, and (5) 3D printing of biomodel. Image acquisition and denoising
Fig. 1. Radiography to 3D graphy – the improvisation pyramid for surgical planning and simulation.
simulate the procedure and if required build customized templates and jigs based on the patients disease, anatomy and his preferences [4,5]. While there are isolated case reports, case series, technical notes and review articles on the subject, there are no systemic studies to analyze the impact of these models on surgeons perception and ultimate surgical execution using a structured reporting template. We used 3D printed models (3D POST1) to plan and simulate the procedure in 50 cases of complex periarticular trauma, complex primary and revision arthroplasty. As a part of the study, benefits and the utility of 3D POST was assessed by using a structured surgeon reported templates. Aims 1. Prospective evaluation of surgical cases performed using 3D printed biomodels. 2. To develop guidelines for choosing appropriate “open source” computing algorithms, rapid prototyping technique and raw materials for creating 3D printed biomodels.
Method Study group The study comprised of a consecutive series of 50 patients, for whom the operating surgeon used a 3D printed biomodel for surgical planning and execution. The study was inclusive of all types of orthopaedic surgeries. (Table 1).
Table 1 Number of cases in the series. Region
No. of cases
Acetabulum Proximal humerus Elbow (periarticular) Neck femur Tibial condyle Foot and ankle Periarticular miscellaneous Complex primary hip arthroplasty Revision hip arthroplasty Peri-prosthetic fractures Total
11 4 4 3 6 4 3 7 6 2 50
CT scan was used as the imaging modality in all cases as it best captures bone data. The CT scan was done using the MRCP (Medical Rapid-prototyping Computed-tomography Protocol) guidelines (Table 2) as described earlier by our group, in 38 (76%) cases [6]. The protocol could not be standardized for the rest, as the CT scan was already performed before the surgeon requested for a biomodel to be made. In brief, the MRCP protocol lays down guidelines for CT image acquisition such that is maximizes the resolution of images without significantly increasing the radiation exposure to the patient. CT images were obtained in DICOM format and were transferred either through a cloud server or sent physically on a CD to the model production location. A copy of raw data was always retained as permanent archives for all patients. Denoising: If the images had random noise (especially if previous implant was present), denoising was carried out to prevent metal artefacts in the CAD (computer aided design) models. Thus the first step was sequential denoising of the CT acquired data. We employed the spatial smoothing method, which is a computer algorithm to reduce the noise without loosing anatomical information [7]. Segmentation Segmentation is a process in which the desired area of anatomy is labeled and thus separated from the surrounding undesired structures on axial, sagittal and coronal images. Computer software was used to assign a label to each voxels that was thus selected. We used two methods (software) for generating a model: Osirix [8]: This is a free open source software that has a 3D rendering function called the surface rendering tool. This software has quite limited capability and uses thresholding for segmentation. However, ability to manually modify the segmentation slice by slice does not exist in this software. This was used for the initial 15 cases. As we gained more knowledge about segmentation we switched over to another open-source software developed by Harvard University, namely 3D Slicer. 3D slicer [9]: This software has the capability to use both thresholding and manual segmentation for selection of the desired voxels. For the remaining 35 cases we used this technique to generate the biomodel (Fig. 2). The initial step was to use thresholding to separate high contrast areas such as bone from the surrounding soft tissue. We found that in osteoporotic bones or cancellous bone, a lot of low density bone was not selected. In addition, there were errors due to partial volume effect around intra-articular areas [10]. Therefore, the next step was to manually navigate slice by slice to “paint” the desired areas or “unpaint” unwanted areas of bone. Generation of 3D CAD model After segmentation the model was exported as an STL (Surface Tessellation Language) file. This was subjected to further processing to improve the quality and remove unwanted parts of the 3D model.
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
3
Table 2 MRCP protocol for CT scan image acquisition ideal for creating 3D printed bio models. Parameters
Description
Field of view (FOV) Scout Region of interest (ROI) kV mA Pitch Collimation Slice thickness Slice increment Kernel/algorithm
This is the region of interest. FOV measuring 12 12 inches is adequate Depends on region of interest and helps planning ROI should be identified Automatic Usually automatic 512 512 1.25–1.50 mm 1–1.5 mm 0.625–0.75 (less than 1 mm) Moderate/soft tissue (not to use “bone/detail”)
Fig. 2. Processing of CT acquired date prior to actual printing of 3D Models.
Postprocessing This was done using Meshlab (Visual Computing Lab, Pisa, Italy), Blender (Blender Foundation, Amsterdam, The Netherlands) and Meshmixer (Autodesk Inc, USA). The postprocessing was carried out in the following sequential steps: 1) Down sampling and reducing complexity: The CAD model surface is represented my triangular areas (tessellations) without gaps or overlap. This virtual mesh of triangles can get quite complicated for a complex area, especially with high resolution CT images. This sometimes results in a very large file. Hence, for such files the complexity was reduced to a
manageable level without compromising on the surfaced details of the model. 2) Cleaning: Unwanted segmentation artefacts were removed. In addition, areas of bone that were clinically unimportant were removed to reduce the cost of printing. If any anatomical areas were disjointed, then an artificial bridge was created to connect it to the main structure to preserve the anatomical relationship. Finally, mesh cleaning and healing algorithms were applied to correct any errors. 3) Smoothening: Finally, smoothening algorithm was applied taking care to not lose the anatomic fidelity of the surface (Fig. 3).
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
4
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 3. Postprocessing of STL file (Fig B after smoothening effect).
3D printing of the biomodel
Post surgery assessment
All the models were produced by Timeless Innovation Labs, Mumbai, India using the Wanhao Duplicator Printer with a resolution of 0.1 mm (100 mm). The 3D printing technology used was Fused Deposition Modelling (FDM). FDM produces 3D biomodels by extruding thin strings of molten polymer to form layers as the extruded material hardens instantly post extrusion from the nozzle (Fig. 4) [11]. The polymer used in the study was a thermoplastic resin called ABS (Acrylonitrile Butadiene Styrene). The average time for printing a 6 inch model was 12 h. Once the biomodel was printed, the supports were manually removed and the surface smoothened with a rasp. The models were delivered to the surgeon prior to the surgery. The model was sterilised using Ethylene oxide sterilization
All five surgeons were asked to fill out a standardized questionnaire after the surgery (Fig. 5) where the surgeon rated individual domains within the questionnaire on a scale of 1 to 10 with respect to usefulness of the model. Each biomodel was assessed on 10 criteria and each criterion was given a score between 1 and 10 with 1 being lowest and 10 being highest. Results In all the cases the surgeon found that the model accurately represented the anatomy, helped in preoperative planning and decreased surgical time (Mean score above 8) (Table 2). They also suggested that they would use the model and recommend it to surgeons in future (mean score 8.9). The greatest unhappiness was the time taken for production of model (mean score 4.6). Most surgeons believed that although there may have been some alterations in the way they tackled the cases on based of new information, they did not think that it altered their surgical plan very significantly (mean score 5.7). Rest of the parameters that pertained to inventory management, avoidance of intraoperative complications and postoperative assessment faired above average with mean scores of between 6 and 8 on each parameter (Fig. 6). Following the experience of these cases a streamlined protocol as shown in the flowchart (Fig. 7) was developed for 3D printing biomodels. In contrast to the expensive proprietary software, this protocol was based on freely available open source programs and includes even the fine nuances on how to obtain the appropriate medical imaging for purpose of 3D printing (Table 3). Case examples
Fig. 4. 3D printing of acetabulum being carried out on a Fused deposition modeling (FDM) machine using ABS polymer.
Illustrative case 1: ST a 21 year old case of Juvenile Rheumatoid arthritis had bilateral severe arthritis of the hip with dysplastic changes. Weighing only 21 kg, and being wheelchair bound since age of 6, performing hip arthroplasty was a challenging affair. His
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
5
Fig. 5. Surgeon reported self assessment questionnaire for 3D models.
pelvis was very small and bone very osteoporotic and to compound the problem the anatomy was fairly distorted in the region. (Figure) A CT scan gave an insight on what surgeons could expect during surgery. 3D printed biomodel from this CT Scan post processing (Fig. 8) however allowed the surgeon to simulate the procedure, understand the requirement of a small size cup and order an inventory (Fig. 9). The models proved an invaluable tool in accomplishing a successful surgery. Illustrative case 2: A 25-year-old car accident victim who suffered from a T Type acetabular fracture was scheduled to undergo a surgery. A planning on OSirix was followed by printing the 3D bio model of both injured and mirror image of non injured side. This model gave an accurate visual impression of the fracture pattern (Fig. 10) and also allowed the template of the implant to be contoured according to patients anatomy and fracture pattern using the mirror image model of opposite side (Fig. 11). The surgeon felt that this significantly reduced the surgical time and help accomplish the surgery in desired fashion.
Discussion 3D printing – also known as rapid prototyping, stereolithography, or addititive manufacturing – is ushering in the next technological revolution. These days it is hard not to notice media headlines proclaiming triumphs of surgeons solving complex surgical problems with the aid of 3D printing. Typically newspaper articles will claim that an intrepid surgeon successfully applied recent breakthroughs in 3D printing technology to help his patient. Despite these claims of novelty, it is interesting to know that 3D printing or rapid prototyping technology has been around since the early 1980s [12]. Even a perfunctory glance at the scientific literature in the last decade will reveal how permeative this technology has become, not only in orthopaedic surgery but also in other surgical specialities [13]. In this paper we reveal our journey of developing the applications of this technology for orthopaedic surgery in the last decade (Fig. 12). Traditionally, before the advent of PACS (Picture Archiving and Communication System), the images obtained using advance imaging such as CT and MRI were printed on films losing valuable information in the process. PACS has changed the way we view images adding another dimension to interpretation. With
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
6
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 6. Graphical representation of average score post assessment using the standard questionnaire.
provided by radiologist to generate 3D images as per their specific requirements, allowing them to virtually manipulate these images on their computer screens [14]. Many surgeons feel that the 3D images provided either by the radiology department or reconstructed by themselves using softwares such as Osirix provide almost a true 3D representation of anatomy. However, these 3D computer images are limited by the two-dimensional nature of the computer screen. Depth perception and relative size cannot be comprehended on the computer screen. Real-size 3D printed biomodels are invaluable to study complex anatomy because they can provide that extra dimension. Physical interaction with anatomy, allowing the representative bone to be touched and examined from all angles (tactile stereotaxy) is the true benefit of 3D printed biomodels [13]. In this case series, we have used 3D printed biomodels for complex orthopaedic cases, such as difficult primary and revision arthroplasty, periatricular fractures, and pelvic trauma. The standard of care for these cases has been to study axial images or 2D representation of 3D computer images provided by the radiology department. 3D printed biomodelling for these cases provided the surgeons an opportunity to physically interact (tactile stereotaxy) with the biomodels during the process of preoperative planning. Acetabular fracture surgeries are known to have a steep learning curve. This is even documented in a report by Matta and Merrit where in they showed that for every group of 20 patients operated the experience improved the ability to avoid unsatisfactory reductions and to perform anatomic reductions [15]. The same was echoed by Kebaish, Roy, and Rennie who Table 3 Average scores.
Fig. 7. Flowchart for developing 3D printed models using MRCP protocol and open source free softwares.
expansing of computing prowess, these DICOM images can be easily reconstructed into 3D computer images that allow a more intuitive understanding of pathoanatomy. Technologically savvy orthopaedic surgeons were already manipulating the DICOM data
Parameters
Average score
Accurate representation Turn around time Helpful in preoperative planning Altered the surgical plan Reduced surgical time Improved inventory management Avoiding unforeseen complications Postoperative Xray assessment New information compared to preoperative CT Would recommend its use to other surgeons
8.1 4.6 8.3 5.7 7.8 8.2 6.8 7.6 6.2 8.9
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 8. 3D printed model of pelvis of 21 year old person suffering from Juvenile Rheumatoid arthritis suffering from dysplasia and arthritis of both hips. Given the small size of the pelvis and distorted anatomy, the 3D biomodel proved to be an invaluable tool in planning the surgery.
demonstrated that reductions obtained by less experienced surgeons had significantly lower rates of anatomic reductions [16]. While getting a hands on experience doing live surgeries may be the ideal way of improving ones surgical skills, ability to preoperatively simulate the same by carrying out various surgical steps on an accurate representing 3D printed model of the case is a valuable tool as shown by this study. The study group that comprised of senior level surgeons felt value in performing these simulations and concurred that if given a chance they would recommend to fellow surgeons. In periarticular fracture accurate reduction and implant placement are two key elements. To ensure these, understanding of the anatomy, good intraoperative imaging and optimal spending of the operative time is required. While understanding the fracture pattern is subjective and depends on knowledge and experience, intraoperative fluoroscopy is a two dimensional and entails cumulative radiation exposure risks. It is reported that each minute of exposure (60 shots) is equivalent to one computed radiography exposure or 4rads of radiation and that many peri articular fracture fixation scenarios require many minutes of these exposures [17]. The 3D printed models help understand the
Fig. 9. Simulation being carried out on the 3D model. In this case acetabular reaming was performed and trialing for the cup and stem was done. As was expected, even post reaming the smallest available cemented cup could be implanted.
7
Fig. 10. 3D printed biomodel of a t Type fracture of the acetabulum. Surgeon can visualize the fracture anatomy and decide and design a plan for surgical execution.
complex spatial anatomy of these fractures, help in planning geometric parameters of target plate fixation and avoid wide surgical exposures. The participating surgeons concurred that with the use of these 3D models, the surgical time can be reduced; the amount of bleeding can be decreased, the radiation exposure to surgeon and the operating room team can be reduced and as corollary the chances of wound infection also goes down. While primary total hip replacement has now become a standard procedure well controlled by orthopedic surgeons, there may remain a few primary cases that remain challenging and complex. While dealing with cases such as dysplasia of the hip, arthrodesis and protrusio hip the surgeon have to biomechanics, bone defect and planned reconstructive technique. Similarly in revision scenario, it is important to know the possible cause of failure, strategies implant removal, choose implants and formulate a strategy for adequate reconstruction [18,19]. In both these conditions, having a real sized model of the area helps understand the biomechanics and bone defect better [20]. The surgeons may also be able to carry out simulation by acetabular reaming or femoral rasping to determine the size of the socket and canal. These simulations also provide an insight in the nature of bone quality and strength in a given area. In these studies, we had 10 cases in which the biomodels was used for arthroplasty planning. They proved invaluable and in virtually all cases some alteration in surgical plan was effected because of these 3D printed models. With newer metal printers with capability of printing biocompatible materials now being easily accessible, many companies are
Fig. 11. Printing the mirror image of the opposite hip allowed the surgeon to template and contour the plate to normal anatomy and help restore the injured side as close to its anatomical position as possible.
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058
G Model JINJ 7388 No. of Pages 8
8
V. Bagaria, K. Chaudhary / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
costs reduce [1]. One also has to consider the time and effort spent by the human capital (radiologists, engineers, or the surgeon’s themselves) in producing these biomodels. However, the 3D printer cost has dramatically reduced over the years and a number of open source softwares are available. This has made the technology accessible in poor resource areas such as our country. The ultimate trial would be to determine the value of biomodels which would be determined by considering both the benefits and costs (value = benefit/cost). Conclusion In all the 50 cases the use of 3D printed biomodel was perceived to have reduced the surgical time, improve the understanding of the condition, reduce the inventory, decrease the surgical time and seen as possibly having a favorable impact on patient outcomes. We propose a protocol for using open source software to generation CAD models which then can be printed with the material of choice. References
Fig. 12. A Post operative X rays showing a well reduced fracture and a contoured plate based on printed model.
now 3D printing the wedges and cones and patient specific hip and knee implants may become much more common in coming years [21]. The ultimate benefit of application of a new technology would be measured in terms of patient outcomes. Preoperative planning using 3D biomodels may help reduce surgical time and estimated blood loss. It also has the potential to reduce complications and improve patient outcomes. It is not hard to imagine how it can do so, especially in surgeries for complex pathology. Although our study is an attempt to generate such evidence, a controlled prospective trial would be required to establish the veracity of these claims. Establishing benefit of 3D biomodeling would require consideration of both the case complexity and surgeon experience and this might prove daunting. Models of simple fractures or pathology may have limited benefit. But as complexity of the fracture increases the information gained from the model may exceed that obtained from radiographs or routine CT scans [1]. The benefit provided by 3D modeling is also dependent on the experience of the surgeon. Less experienced surgeons and trainees would find 3D models of moderately complex fractures useful. As surgeon expertise increases the benefit gained from such models would decrease. However, in extremely complex and unusual fracture patterns, even the most experienced surgeons would derive significant benefit from 3D biomodels [1]. The added expense is an important consideration to justify the use of 3D biomodels. Although we did not do a controlled cost analysis in our series, we can predict some assumptions. The initial cost of setting up a 3D printing unit (printers and buying proprietary software) is high, but with time and experience the
[1] Tetsworth KD, Mettyas T. Overview of emerging technology in orthopedic surgery: what is the value in 3D modeling and printing? Tech Orthop 2016. [2] Esses SJ, Berman P, Bloom AI, Sosna J. Clinical applications of physical 3D models derived from mDCT data and created by rapid prototyping. AJR Am J Roentgenol 2011196(6). [3] Sariali E, Boukhelifa N, Catonne Y, Pascal Moussellard H. Comparison of threedimensional planning-assisted and conventional acetabular cup positioning in total hip arthroplasty: a randomized controlled trial. J Bone Joint Surg 2016;98 (2)108–16, doi:http://dx.doi.org/10.2106/JBJS.N.00753 American Volume. [4] Krishnan SP, Dawood A, Richards R, Henckel J, Hart AJ. A review of rapid prototyped surgical guides for patient-specific total knee replacement. J Bone Joint Surg 2012;94(11)1457–61 British volume. [5] Sariali E, Mauprivez R, Khiami F, Pascal-Mousselard H, Catonné Y. Accuracy of the preoperative planning for cementless total hip arthroplasty: a randomised comparison between three-dimensional computerised planning and conventional templating. Orthop Traumatol 2012;98(2):151–8. [6] Bagaria V, Shah S, Bhansali R. Using 3D-printed patient-optimized surgical tools (3D POST) for complex hip and knee arthroplasty N2. In: Bagaria V, editor. Arthroplasty – A Comprehensive Review. Rijeka: InTech; 2016 Ch. 09. [7] Thilagamani S, Shanthi N. Gaussian and gabor filter approach for object segmentation. J Comput Inf Sci Eng 2014;14(2):021006. [8] Rosset A, Spadola L, OsiriX Ratib O. An open-source software for navigating in multidimensional dICOM images. J Digit Imaging 2004;17(3):205–16. [9] Fedorov A, Beichel R, Kalpathy-Cramer J, et al. 3D slicer as an image computing platform for the quantitative imaging network. Magn Reson Imaging 2012;30 (9):1323–41. [10] Green N, Glatt V, Tetsworth K, Wilson LJ, Grant CA. A practical guide to image processing in the creation of 3D models for orthopedics. Tech Orthop 2016;31 (3):153–63. [11] Gür Y. Additive manufacturing of anatomical models from computed tomography scan data. Mol Cell Biomech MCB 2014;11(4):249–58. [12] Hull C. Apparatus for production of three-dimensional objects by . . . 1996, http://www.google.com/patents/US5556590. [13] Martelli N, Serrano C, van den Brink H, et al. Advantages and disadvantages of 3-dimensional printing in surgery: a systematic review. Surgery 2016;159 (6):1485–500. [14] Reddix RN, Webb LX. Computer-assisted preoperative planning in the surgical treatment of acetabular fractures. J Surg Orthop Adv 2007;16(3):138–43. [15] Matta JM, Merritt PO. Displaced acetabular fractures. Clin Orthop Relat Res 1988;230:83–97. [16] Kebaish AS, Roy A, Rennie W. Displaced acetabular fractures: long-term follow-up. J Trauma 1991;31(11):1539–42. [17] Lo NN, Goh PS, Khong KS. Radiation dosage from use of the image intensifier in orthopaedic surgery. Singapore Med J 1996;37(1):69–71. [18] Robertson DD, Sutherland CJ, Lopes T, Yuan J. Preoperative description of severe acetabular defects caused by failed total hip replacement. J Comput Assist Tomogr 1998;22(3):444–9. [19] Munjal S, Leopold SS, Kornreich D, Shott S. Finn HA: CT-generated 3dimensional models for complex acetabular reconstruction. J Arthroplasty 2000;15(5):644–53. [20] Yu AW, Duncan JM, Daurka JS, Lewis A, Cobb J. A feasibility study into the use of three-dimensional printer modelling in acetabular fracture surgery. Adv Orthop 20152015(4) 617046-4. [21] Wong KC, Kumta SM, Geel NV, Demol J. One-step reconstruction with a 3Dprinted, biomechanically evaluated custom implant after complex pelvic tumor resection. Comput Aided Surg 2015;20(1):14–23.
Please cite this article in press as: V. Bagaria, K. Chaudhary, A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.08.058