Abstract
There has been a recent emphasis in surgical science on supplementing surgical training outside of the Operating Room (OR). Combining simulation training with the current surgical apprenticeship enhances surgical skills in the OR, without increasing the time spent in the OR practicing. Computer-assisted surgical (CAS) planning consists of performing operative techniques virtually using three-dimensional (3D) computer-based models reconstructed from 3D cross-sectional imaging. The purpose of this paper is to present a CAS system to rehearse, visualize and quantify osteotomies, and demonstrate its usefulness in two different osteotomy surgical procedures, cranial vault reconstruction and femoral osteotomy. We found that the system could sufficiently simulate these two procedures. Our system takes advantage of the high-quality visualizations possible with 3DSlicer, as well as implements new infrastructure to allow for direct 3D interaction (cutting and positioning) with the bone models. We see the proposed osteotomy planner tool evolving towards incorporating different cutting templates to help depict several surgical scenarios, help ‘trained’ surgeons maintain operating skills, help rehearse a surgical sequence before heading to the OR, or even to help surgical planning for specific patient cases.
Keywords: Osteotomy, Surgical planning, 3D Slicer, Craniosynostosis
1. INTRODUCTION
Currently, surgical science is characterized by rapid changes in knowledge, procedures, and technology1. The approaches that supplement surgical training outside of the Operating Room (OR) have increased in recent years as OR time has become too valuable to permit refinement of a rapidly expanding set of surgical skills. A combination of simulation training with the current surgical apprenticeship enhances surgical skills in the OR. Computer-assisted surgical (CAS) planning consists of performing operative techniques virtually using three-dimensional (3D) computer-based models reconstructed from 3D cross-sectional imaging. CAS planning offers impressive 3D views of the existing anatomy in the surgical site, allows rehearsing surgical steps or surgical techniques in the 3D models, provides quantitative metrics of surgical outcomes and overall adds a new perspective for the surgeon performing the procedure. These technologies augment the ability of an experienced surgeon to perform certain types of surgery before entering the OR. Quantitative tools based on 3D models provide an objective assessment of the anatomy that can help the surgeon quantify the existing pathology, plan the surgical steps to perform, and potentially reduce peri-operative time. In this paper, we present our work developing a CAS planning tool for osteotomy simulation for two scenarios: femoral osteotomy and cranial vault reconstruction. An osteotomy is a surgical procedure that involves cutting one of several bony structures to correct pathology, deformations, alignment or position. Depending on the osteotomy site, recovery can be extensive and lengthy for the patient.
The purpose of this paper is to present a CAS system designed to rehearse, visualize and quantify osteotomies, as well as to demonstrate its usefulness in a number of surgical areas. CAS can help with overall surgical performance, increase interventional, decrease pre-operative time, and this improve patient outcome.
1. MATERIALS
We validate and present results from our CAS software system using two osteotomy cases:
2.1. Craniosynostosis case
Craniosynostosis is a condition in which one or more of the fibrous skull sutures in a young infant prematurely fuse. These children may require a full cranial vault reconstruction surgery to correct and prevent head shape malformations. For example, the surgery for repairing coronal synostosis (fusion of the suture on the side of the skull) aims to widen the skull regions affected by the fusion and abnormal growth so that the brain can grow to a more normal volume/shape on its own. We used computed tomography scans of an 8-month-old infant with a diagnosis of coronal synostosis to illustrate the usability of the CAS module. Skull segmentations for this data were produced using an established image analysis pipeline for craniosynostosis cases2–4.
1.2. Femoral Osteotomy
Femoral osteotomy is a surgical procedure that is performed to correct specific deformities of the femur. For demonstration, we used the BodyParts3D database5. BodyParts3D is a dictionary-type database for anatomy in which anatomical concepts are represented by 3D models that specify segments of a 3D whole-body model for an adult human male. It encompasses morphological and geometrical knowledge in anatomy and complements ontological representation. BodyParts3Dis licensed under CC Attribution-Share Alike 2.1 Japan. We simulated a femoral osteotomy using a femur model obtained from this database.
2. METHODS
The emergence of open source libraries and tools in the last decade has changed the process of academic software development and continues to contribute to the free exchange of information and methods. Recently, increasing resource sharing and thus reproducibility has become one of the focus initiatives of the National Institutes of Health6. These open access or open source developments have contributed significantly to providing small research labs the unprecedented ability to generate considerable contributions to the field based on tools such as 3DSlicer7. We decided to develop and aim to disseminate our Osteotomy Planner using 3DSlicer.
2.1. Osteotomy Planner Software Design
Figure 1 shows the Osteotomy Planner GUI and software workflow. The bone structure is modeled as a group of 3D mesh models with each individual model representing a freely movable piece. Each bone in the group can be rotated and translated in 3D space to change its position relative to the rest of the bones, and can be re-colored for easy differentiation. The bones can then be cut and bent to adjust the surface area and enclosed volume. After performing adjustments, the Osteotomy Planner can generate a number of metrics for analyzing changes to bone structure.
Figure 1.
a) Osteotomy Planner 3DSlicer Extension Graphic User Interface (GUI), displaying all the functional units in the architectural diagram. b) Architecture of the Osteotomy Planner 3DSlicer Extension. The module is coded as a C++ loadable module, and uses C++ CLIs to implement discrete tasks.
Metrics are generated based on the current bone structure alone, or by comparison of the bone structure to user defined reference models or the previous state of the structure.
2.2. Osteotomy Planner Task Breakdown
2.2.1. Cutting
Cutting is implemented using arbitrarily placed and oriented planes (see Figure 2). The user can place the cutting plane and preview the resulting bone split. Once the cut is finalized, two new bones (the result of the split) are added to the structure group and the original bone is removed. Bones may be cut multiple times by repeating the cut procedure on the new bones.
Figure 2.
Simulated orbital rim advancement procedure: a) pre-surgical craniosynostosis skull b) placing the plane to separate the frontal bone c) advancing the frontal bone to increase the enclosed volume
2.2.2. Bending
Bending actions are performed after bone cutting is completed. Bending is implemented using a simple hinge model and a thin plate spline for interpolation (see Figure 3). The user selects four control points, two which defined the axis of bending, and two to define the direction (Figure 3a). The user can further specify whether they want to bend on both sides of the axis, or only on a single side. A slider is used to control the magnitude of the bending action.
Figure 3.
Bending procedure for craniosynostosis case: a) User defined control points. Green line shows the bending direction, red line shows the bending axis b) Side view of unbent bone c) bone bent outward d) bone bent inward.
2.2.3. Metrics
Currently, two main metrics are calculated for each bone model: volume and surface area. The entire bone structure can also be merged using the vtkAppendFilter into a single model with metrics calculated for the merged model. The merged bone model can also be “shrink-wrapped” using the vtkSmoothPolyDataFilter to provide an estimate of enclosed volume, for cases where that is relevant.
The Osteotomy Planner also allows the user to load reference models for computing additional metrics. The two main metrics can be computed for these reference models for comparison to the post-operative metrics. The planner currently only supports reference models that represent the merged bone group (i.e. not each individual piece). The planner can also generate a “pre-operative” reference model from the bone group before any changes are made. Model to model distances can also be computed between the bone structure and the references. These distances can be displayed on the bone models (see Figure 4).
Figure 4.
Distance metrics for craniosynostosis case. Absolute distance between the bone model and the brain reference model is show using heat map coloring on the bone surface. Hotter areas indicate larger distance between the bone and the reference.
2.3. Osteotomy Planner Software Implementation
The tool architecture uses 3DSlicer specific C++ libraries, as well as VTK8 C++ libraries (see figure 1b). All the source code is freely available via GitHub9.
The Osteotomy Planner makes use of a key new feature in 3D Slicer that allows 3D free-form interactive transformations thanks to a special visualization widget visible in the 3D view of the program. This enhancement allows the user to manipulate the transform applied to an object by clicking and dragging control points on the transform object in the 3D view. Rotations, translations and rescaling can be applied in this way. The interactive transform is used to implement the bone re-arrangement. The Planner also uses a custom 3DSlicer fiducial type (the vtkMRMLPlanesMarkupNode) to control the cutting task. This fiducial can be placed and oriented to create a cutting plane (Figure 2b).
3. RESULTS
3.1. Frontal advancement osteotomy procedure
Using the craniosynostosis data set, we simulated an orbital rim advancement (Figure 2). In this procedure, the fused frontal bone is separated from the rest of the skull, and moved forward to increase the enclosed volume. Figure 2b shows the placement of the plane to separate the frontal bone, shown in green. Figure 3 shows bending actions applied to the craniosynostosis case. Control points are placed to define the direction and axis of bending, and typical inward and outward bends are shown.
3.2. Femoral osteotomy procedure
Using the femur model, we simulated a femoral osteotomy (see Figure 5). The purpose of this procedure is to correct the angle between the ball joint and the long axis of the femur. The model used was of a healthy femur, so the procedure results in an over-corrected joint angle. First, the joint is separated from the femur (Figure 5b). A second cut is then made to create a new attachment surface for the joint (Figure 5c). Finally, the joint is repositioned at the corrected angle (Figure 5d).
Figure 5.
Simulated femoral osteotomy: a) original model b) creating the first cut to separate the ball joint c) placing the plane for the second cut to create a new attachment surface for the joint d) reattachment of the ball joint at a new angle.
4. DISCUSSION
In this paper, we presented a tool designed to help junior surgeons in training or provide a tool for experienced surgeons to simulate and optimize specialized types of surgery before it is time to go into the OR. The proposed CAS tool will allow surgeons to visualize osteotomies in a virtual environment and obtain quantitative metrics about the morphology of the bone before and after the osteotomy. We believe this will help surgeons contemplate the outcomes of interventional decisions before actual surgery. More importantly, the methods used in this study are disseminated and included in an open source image analysis toolbox. Commercial software packages produce adequate surface reconstructions and/or offer landmark, surface and/or voxel-based registration methods, but they are not open source, cannot be modified, do not interact well with each other and do not provide flexibility for customization. Due to its open licensing, 3D Slicer represents the perfect disseminating vehicle for our Osteotomy planner extension.
We see the proposed osteotomy planner tool evolving towards incorporating different cutting templates to help depict several surgical scenarios, help ‘trained’ surgeons maintain operating skills, rehearse a surgical sequence before heading to the OR, and optimize surgical planning for specific cases to improve patient outcome. Future steps will look into validating this tool clinically. We will investigate quantifying the improvements in surgery after using the proposed Osteotomy Planner.
ACKNOWLEDGEMENTS
This work was supported by the National Institute of Health (NIH) Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) grant 5R42HD081712-03 (Image-guided Planning System for Skull Correction in Children with Craniosynostosis: Phase II).
REFERENCES
- [1].V Polavarapu H, Kulaylat AN, Sun S, and Hamed OH, “100 years of surgical education: the past, present, and future.,” Bull. Am. Coll. Surg, vol. 98, no. 7, pp. 22–7, Jul. 2013. [PubMed] [Google Scholar]
- [2].Mendoza CS, Safdar N, Myers E, Kittisarapong T, Rogers GF, and Linguraru MG, “Computer-Based Quantitative Assessment of Skull Morphology for Craniosynostosis,” Springer, Berlin, Heidelberg, 2013, pp. 98–105. [Google Scholar]
- [3].Mendoza CS, Safdar N, Okada K, Myers E, Rogers GF, and Linguraru MG, “Personalized assessment of craniosynostosis via statistical shape modeling,” Med. Image Anal, vol. 18, no. 4, pp. 635–646, May 2014. [DOI] [PubMed] [Google Scholar]
- [4].Porras AR, Zukic D, Equobahrie A, Rogers GF, and Linguraru MG, “Personalized optimal planning for the surgical correction of metopic craniosynostosis,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2016, vol. 9958 LNCS, pp. 60–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Mitsuhashi N, Fujieda K, Tamura T, Kawamoto S, Takagi T, and Okubo K, “BodyParts3D: 3D structure database for anatomical concepts,” Nucleic Acids Res, vol. 37, no. SUPPL. 1, pp. D782–D785, Jan. 2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Baker M, “1,500 Scientists Lift the Lid on Reproducibility,” Nature, vol. 533, no. 7604, pp. 452–454, 2016. [DOI] [PubMed] [Google Scholar]
- [7].“3DSlicer.” [Online]. Available: www.slicer.org.
- [8].Schroeder W, Martin K, and Lorensen B, Visualization Toolkit: An Object-Oriented Approach to 3D Graphics, 4th Edition. Kitware, 2006. [Google Scholar]
- [9].Medical K, “Osteotomy Planner Github Repository,” 2017. [Online]. Available: https://github.com/KitwareMedical/OsteotomyPlanner.