Design of a 3D Model for Temporal Bone Surgical Simulation Austin S. Rose, MD1, Julia S. Kimbell, PhD1, Caroline E. Webster2, Ola L.A. Harrysson, PhD2, Eric J. Formeister, MS1 and Craig A. Buchman, MD1 1 2
Department of Otolaryngology-Head & Neck Surgery, University of North Carolina School of Medicine Department of Industrial and Systems Engineering, North Carolina State University
Julia S. Kimbell, PhD Department of Otolaryngology - Head & Neck Surgery University of North Carolina School of Medicine, CB #7070 Chapel Hill, NC 27599-7070
[email protected] Caroline Webster Edward P. Fitts Department of Industrial and Systems Engineering 414-D Daniels Hall, CB#7906 Raleigh, NC 27695-7906
[email protected] Ola L.A. Harrysson, PhD Edward P. Fitts Department of Industrial and Systems Engineering 414-D Daniels Hall, CB#7906 Raleigh, NC 27695-7906
[email protected] Eric J. Formeister Department of Otolaryngology - Head & Neck Surgery University of North Carolina School of Medicine, CB #7070 Chapel Hill, NC 27599-7070
[email protected] Craig A. Buchman, MD Department of Otolaryngology - Head & Neck Surgery University of North Carolina School of Medicine, CB #7070 Chapel Hill, NC 27599-7070
[email protected]
Corresponding Author: Austin S. Rose, MD Department of Otolaryngology - Head & Neck Surgery University of North Carolina School of Medicine, CB #7070 Chapel Hill, NC 27599-7070
[email protected] 919-966-3342 (phone) 919-966-7941 (fax)
Financial disclosures / conflict of interest: The authors have nothing to disclose. Keywords: 3D printing, additive manufacturing, temporal bone, simulation
ABSTRACT As the process of additive manufacturing, or three-dimensional (3D) printing, has become more practical and affordable, a number of applications for the technology in the field of Otolaryngology-Head & Neck Surgery have been considered. One area of promise is temporal bone surgical simulation. In this project, 3D models of human temporal bones were created from de-identified temporal bone computed tomography (CT) scans using biomedical image processing software and ultimately printed in a variety of materials. The various models created were dissected in a temporal bone laboratory and graded by attending otolaryngologists, using a 5-point Likert scale, on their anatomical accuracy and suitability as a simulation of cadaveric and operative temporal bone drilling. Simulated temporal bones created by this process may be of potential benefit in surgical training, pre-operative simulation for difficult otologic cases, and in the standardized testing of temporal bone surgical skills.
INTRODUCTION Simulation in medical and surgical education is widely recognized as an important tool in preparing physicians for direct patient care, in protecting patients that undergo surgery and in reducing medical errors. Ideally, clinical simulations as a part of training allow for the learning and development of skills in a realistic, yet safe environment. The practice of clinical scenarios is often accomplished using simulated, standardized patients portrayed by actors. For surgical procedures however, simulation has traditionally made use of physical models, animals or human cadavers. More recently, computer-based simulation systems have been developed, allowing for numerous virtual reality (VR) clinical situations and the easy integration of didactic material. Hybrid devices also exist, incorporating physical models or hand-held instruments for a more realistic interface with the computer. Simulators of this type are already in use by training programs in general surgery, neurosurgery, ophthalmology and orthopedic surgery, as well as in otolaryngology.1 It has been suggested that surgical simulators may soon be considered advanced enough for use in standardized testing and certification.2 The goal for any surgical simulation should be to realistically reproduce the visual, auditory, haptic and other aspects that are characteristic of the actual procedure. The incorporation of new technologies in medical simulation has allowed for substantial advances in this regard. The result will no doubt be better training for surgeons and significant improvements in patient safety.
Improvements in surgical simulation with new technologies may prove timely, especially as a number of changes affecting resident training raise concerns for decreased patient contact time, surgical volume and experience. Recent ACGME-required reforms, including restrictions on duty hours in particular, though shown to reduce medical errors, may be having a negative impact on the surgical confidence and competence of graduating residents.3,4,5 Increased surgical simulation during training may prove an important part of the flexible and creative solutions necessary to remedy this problem. The field of otolaryngology is particularly well-suited to the integration of clinical simulation as a part of its training. Models, cadavers, animals and the use of 3D computer simulations have all been used in teaching complex anatomy and in simulating critical surgical procedures, including tracheotomy, bronchoscopy, endoscopic sinus surgery and temporal bone dissection.1,2 The use of both fixed and fresh human cadaver temporal bones in labs specially outfitted for temporal bone drilling has long been a core component in training for otologic surgery, though cadaver temporal bone dissection has certain limitations. Even fixed human cadaver specimens may potentially carry infectious agents such as Mycobacterium tuberculosis, Hepatitis B and C viruses, and prions associated with encephalopathies such as Creutzfeldt-Jakob disease.6,7 Exposure to formaldehyde itself, when used for fixation, may be a potential carcinogen.8 In addition, injuries to trainees (such as a significant laceration of the hand at our institution in 2012) are of concern during the harvesting of temporal bones from cadavers and the removal of soft tissue in preparation for dissection. The cost of human temporal bones can also vary widely and may be prohibitive in some countries where specimens are in short supply due to financial or cultural issues. In one study from Pakistan in 2011, 37%
of medical students avoided cadaveric dissection due to moral or ethical grounds and another 18.6% for religious reasons.9 An additional drawback in using cadaver temporal bones for training is the inability to either standardize specimens for the development and testing of basic skills, or to provide customized scenarios for teaching of difficult or unusual cases. This is addressed somewhat by VR temporal bone simulators, though with some sacrifice of a realistic drilling and operative experience.10 To address the clear need for improved simulations in otolaryngology education, and of temporal bone dissection in particular, we developed a detailed, physical temporal bone model using additive manufacturing for use in conventional temporal bone dissection laboratories. Additive manufacturing is the process of creating 3D objects by adding successive layers of material in different shapes. It is particularly useful in the creation of anatomical models with highly detailed internal anatomy such as the temporal bone. Recently developed 3D printers can also print in combinations of materials and colors, allowing for simulation of not only bone, but of adjacent soft tissue and neurovascular structures as well. Several challenges exist in creating a simulated temporal bone that is comparable to cadaver specimens or actual experience in the operating room (OR). Of primary importance is the ability to adequately simulate the highly detailed bony and soft tissue anatomy characteristic of the human temporal bone. An additional challenge is the need for materials that acceptably re-create the experience of drilling through both cortical and trabecular bone, as well as simulate the identification and dissection of important soft tissue structures including the facial nerve, sigmoid sinus and internal carotid artery. Our
hypothesis is that a simulated temporal bone can be created using 3D printing that will prove both safe and beneficial in training for actual temporal bone surgical cases.
MATERIALS AND METHODS Model Development Institutional review board approval for the use of de-identified imaging data was obtained prior to the study. High-resolution (0.3 mm slices) CT scans of human temporal bones were selected for use in the study. Standard digital imaging and communications in medicine (DICOM) data from the scans were then imported into specialized software in which the bony anatomy was segmented from surrounding structures and stereolithography (STL) files appropriate for printing were created. STL files use triangulation to describe the surface geometry of a three-dimensional object. These files were then used for the physical 3D printing of temporal bone models in a variety of materials using the stereo-lithography process (SLA) for additive manufacturing (see Figure 1). As no additional image processing had been performed at this point beyond the segmentation of bone, the first two models represented only bony anatomy. The first model (A) was produced using VeroWhite resin as the printing material on an Objet350 Connex printer (Stratasys, Eden Prairie, MN). A second model (B) was also produced using non-proprietary nylon-infused plaster. Models A & B were then tested in a temporal bone laboratory using a Stryker Core electric otologic drill (Stryker, Kalamazoo, MI) with both cutting and diamond burrs. Following this round of test drilling, it became apparent that more could be done to improve the level of anatomic detail of the models. In addition, the Department of
Industrial and Systems Engineering at North Carolina State University (Raleigh, NC) was contacted for input regarding more realistic materials. A third, more detailed, model (C) was then produced. Prior to printing, the DICOM data were first imported into commercial image-processing software (Mimics, Materialise, Belgium), which allowed for further refinement of bone anatomy and the additional segmentation of soft tissue structures, including the facial nerve, greater superficial petrosal nerve, tympanic membrane, sigmoid sinus and internal carotid artery (see Figure 2). The model was then fabricated using a combination of proprietary thermoset polymers on an Objet350 Connex printer. Varying ratios of materials were used to achieve unique biomechanical properties for different anatomical structures.
Validation Study While the pursuit of an ideal simulated temporal bone is an ongoing process of development and refinement, initial testing of Model C was promising enough to conduct a face validity study. Eight attending otolaryngologists were recruited to participate in the dissection of simulated temporal bones (Model C), after which they completed a Likert scale questionnaire (see Figure 3). Their post-graduate year following medical school graduation ranged from 7 to 39 years (average = 17.6 years). The first five questions asked subjects to rate different aspects of their experience in comparison to drilling a fixed, human cadaver temporal bone (1 = unlike to 5 = identical). Additional questions addressed ease of use, safety, possible irritation due to simulated bone dust and the overall value of the 3D printed temporal bone as a surgical simulation in preparation for cases in the OR. The dissection of the models was
performed in a temporal bone laboratory and all participants used gowns, gloves, masks and eye protection during the study. A brief summary of the 3D printing process used, including an explanation of support material, was given to all participants prior to the dissection.
RESULTS Models A and B were dissected as described above by the principle investigator. Model A, printed in VeroWhite resin, demonstrated characteristics that might be expected with a plastic, such as nylon, when drilled. In particular, plastic-like shavings were generated and the overall drilling experience was unlike that of a cadaver temporal bone. Anatomic detail was also lacking. Though an aerated mastoid antrum could be identified, there was little additional definition of mastoid air cells or other structures. Model B, printed with nylon-infused plaster, demonstrated improved aeration of some mastoid air cells and drilling characteristics more comparable to those of cadaver temporal bones. Still, though more similar to bone than Model A, the material remained too soft and cutting burrs passed through it far too quickly to simulate actual bone. As described above, Model C, which incorporated a range of color and material properties, represented a significant improvement in the overall simulated temporal bone model. Dissection revealed detailed air cells of the mastoid, and allowed for easy identification of the sigmoid sinus, middle fossa plate, horizontal canal, facial nerve, facial recess and ossicles (see Figure 4). As the primary investigators felt this model had reached a level of probable significant value in surgical training, it was used to conduct a face validity study.
After completion of the questionnaire by all eight attending otolaryngologists, the statistical mean for each of the nine Likert scale questions was then calculated and a box plot generated (see Figure 5). In response to the single yes or no question, none of the participants reported irritation of the skin, eyes, lungs or mucous membranes. The average rating for the value of the simulated temporal bone in surgical experience and training in comparison to cadaveric bones was 4.2. High ratings were also given for ease of use (4.8) and safety (4.7). Bony and soft tissue anatomy received average ratings of 3.9 and 4.0 respectively, while the likeness of drilling dissection for cortical bone and trabecular bone were both rated at 4.0. For its value in surgical experience and training in comparison to temporal bone drilling in the OR, the simulated temporal bone was rated 3.8 on average. The average rating for overall value as a surgical simulation in preparation for cases in the OR was 4.4.
DISCUSSION Developing a temporal bone simulation comparable to the OR experience, or even the dissection of cadaver specimens, involves a constant process of refinement. In early attempts, physical models of the temporal bone were built layer by layer using laminated paper as the printing material. By 2007, Suzuki et al produced models of the temporal bone and inner-ear, using polyamide nylon and glass beads, that were useful primarily for teaching, though the authors discussed their potential as a substitute for cadaver specimens in temporal bone surgical training even at that time.11 In recent years, the capabilities of 3D printing technology have rapidly improved, while associated costs have dropped significantly. This has helped various groups to
address some of the challenges inherent in creating a successful 3D model for temporal bone surgical simulation, including improvements in anatomic detail and the use of more realistic materials. In 2010, Bakhos et al used white resin to reproduce the bony anatomy of the temporal bone based on CT scans of cadaveric specimens. Ossicular detail was sufficient for the simulation of middle ear prosthesis placement in two of the models.12 In 2013, Mick et al developed temporal bone prototypes with an improved similarity to bone, and in multiple colors, using plaster powder and a binding agent containing cyanoacrylate. The middle fossa plate was coated with latex paint to simulate dura.13 Recently, Hochman et al used a novel slicing algorithm to address the problem of support material filling void spaces within the model during printing.14 However, in our study, neither the primary investigators nor the attending participants found support material, within the mastoid air cells for example, to significantly hinder the simulation. Hochman’s group also studied the mechanical properties of their material, a proprietary composite powder, for its likeness to bone in terms of hardness, vibration, acoustic properties, drill skip, visual characteristics and overall similarity. Four different binding agents for infiltration of the powder material were used, and cyanoacrylate with hydroquinone performed the best. As work towards an ever more realistic 3D model of the human temporal bone continues, the potential benefits for education and training become clear. Threedimensional modeling and printing allow for both standardization in teaching and evaluation, as well as for customized simulations. A digital library of both adult and pediatric temporal bones with unusual anatomy or various pathologic scenarios could be developed and printed as needed for simulation in the temporal bone laboratory.
Additionally, temporal bone models of this type could be rapidly produced for preoperative simulation of specific challenging cases, perhaps helping to avoid any unforeseen pitfalls during the actual operation. Though further study is required, the potential for better patient outcomes and safety, as a result of improved skills and fewer surgical errors, seems promising. A safer experience for trainees is another area of benefit, as no physical preparation is required before drilling the 3D model. Lastly, there is a possibility in some cases for decreased costs and increased availability of 3D models in comparison to cadaver temporal bone specimens. Though significant progress has been made, a number of challenges in producing a reasonably accurate temporal bone simulation remain. Advances in anatomic detail will depend on improvements in both the resolution of input data as well as subsequent image processing. The increasing availability of 3D printers that produce objects in multiple colors and materials should enhance our ability to accurately represent both bone and soft tissue structures. In addition, one can foresee advances in materials that allow for compatibility with intra-operative facial nerve monitoring as a part of the simulation. Our work thus far has demonstrated the ability to produce a 3D temporal bone simulation with significant anatomic detail and a likeness to human cadaver specimens for drilling and dissection. Our model is among the first to blend multiple materials in the printing process, achieving unique colors and physical characteristics for different anatomical structures. In addition, our use of DICOM data directly from a clinical CT scan shows the potential to rapidly produce 3D temporal bone models upon request for pre-operative surgical simulation.
CONCLUSION The results of this study support the hypothesis that a simulated 3D temporal bone can be created with significant value in otologic training and education, while remaining safe and easy to use. While other approaches exist, including injection-molded plastic models, cadaveric dissection and computer-based VR simulations, the use of 3D design and printing offers a number of specific advantages over more traditional modes of surgical simulation.
REFERENCES 1.
Khemani S, Arora A, Singh A, Tolley N, Darzi A. Objective skills assessment and construct validation of a virtual reality temporal bone simulator. Otol Neurotol 2012; 33(7):1225-1231
2.
Abou-Elhamd KA, Al-Sultan AI, Rashad UM. Simulation in ENT medical education. J Laryngol Otol 2010; 124:237-241
3.
Landrigan CP, Rothschild JM, Cronin JW et al. Effect of reducing interns’ work hours on serious medical errors in intensive care units. N Engl J Med 2004; 351(18):1838-1848
4.
Fonseca AL, Reddy V, Longo WE, Udelsman R, Gusberg RJ. Operative confidence of graduation surgery residents: a training challenge in a changing environment. Am J Surg 2014; [Epub ahead of print]
5.
Ahmed N, Devitt KS, Keshet I et al. A systematic review of the affects of resident duty hour restrictions in surgery: impact on resident wellness, training and patient outcomes. Ann Surg 2014; [Epub ahead of print]
6.
Demiryurek D, Bayramoglu A, Ustacelebi S. Infective agents in fixed human cadavers: a brief review and suggested guidelines. Anat Rec (New Anat) 2002; 269:194-197
7.
Correia JC, Steyl JL, De Villiers HC. Assessing the survival of Mycobacterium tuberculosis in unembalmed and embalmed human remains. Clin Anat 2014; 27(3):304-307
8.
Whitehead MC and Savoia MC. Evaluation of methods to reduce formaldehyde levels of cadavers in the dissection laboratory. Clin Anat 2008; 21(1):75-81
9.
Nas S, Nazir G, Iram S et al. Perceptions of cadaveric dissection in anatomy teaching. J Ayub Med Coll Abbottabad 2011;23(3):145-148
10.
Arora A, Swords C, Khemani S et al. Virtual reality case-specific rehearsal in temporal bone surgery: a preliminary evaluation. Int J Surg 2014; 12(2):141145
11.
Suzuki M, Hagiwara A, Ogawa Y, Ono H. Rapid-prototyped temporal bone and inner-ear models replicated by adjusting computed tomography thresholds. J Laryngol Otol 2007; 121:1025-1028
12.
Bakhos D, Velut S, Robier A, Al zahrani M and Lescanne E. Threedimensional modeling of the temporal bone for surgical training. Otol Neurotol 2010; 31:328-334
13.
Mick PT, Arnoldner C, Mainprize JG, Symons SP, Chen JM. Face validity study of an artificial temporal bone for simulation surgery. Otol Neurotol 2013; 34(7):1305-1310
14.
Hochman JB, Kraut J, Kazmerik K, Unger BJ. Generation of a 3D printed temporal bone model with internal fidelity and validation of the mechanical construct. Otolaryngol Head Neck Surg 2014; 150(3):448-454
LEGEND Figure 1. Simulated temporal bones printed in VeroWhite (Model A), nylon-infused plaster (Model B) and multiple materials (Model C – lateral and medial aspects). Lateral aspect of simulated temporal bone with external auditory canal (EAC) and zygomatic arch (ZA). Medial aspect of simulated temporal bone with internal carotid artery (ICA), facial nerve (FN) in the internal auditory canal and sigmoid sinus (SS).
Figure 2. DICOM data imported into 3D image-processing software, allowing the segmentation bone and creation of of soft tissue structures.
Figure 3. Likert scale questionnaire used in face validity study.
Figure 4 (clockwise from upper left). Dissection of simulated temporal bone in standard temporal bone laboratory. Malleus and incus seen through the external auditory canal in simulated temporal bone. Mastoid antrum (MA) and facial recess (FR) following dissection of simulated temporal bone; the facial nerve (FN) and sigmoid sinus (SS) are also seen. Human cadaver temporal bone following dissection for comparison
Figure 5. Results of Likert scale questionnaire.
FIGURES Figure 1 (A, B and C)
Figure 2
Figure 3 Please circle a response for each question on a scale of 1 to 5 ( 1=unlike, 5 = identical ): 1.
Please rate the bony anatomy detail in comparison to drilling a cadaveric temporal bone. 1
-
unlike
2.
-
2
-
very different
2.5
-
3
similar
-
3.5
-
4
-
very similar
4.5
-
5
identical
Please rate the soft tissue anatomy you encountered (sigmoid sinus, facial nerve, carotid artery) in comparison to drilling a cadaveric temporal bone. 1
-
unlike
3.
1.5
1.5
-
2
-
very different
2.5
-
3
similar
-
3.5
-
4
-
very similar
4.5
-
5
identical
Please rate the likeness of drilling dissection (haptic feel, visual experience, sound) in comparison to a cadaveric temporal bone for the cortical bone. 1
-
unlike
1.5
-
2
-
very different
2.5
-
3
similar
-
3.5
-
4
-
very similar
4.5
-
5
identical
4.
Please rate the likeness of drilling dissection (haptic feel, visual experience, sound) in comparison to a cadaveric temporal bone for the trabecular bone. 1
-
unlike
5.
1.5
-
2
-
very different
2.5
-
3
similar
-
3.5
-
4
-
very similar
4.5
-
5
identical
Please rate the printed temporal bone for its value in surgical experience and training in comparison to a cadaveric temporal bone. 1
-
unlike
1.5
-
2
-
very different
2.5
-
3
similar
-
3.5
-
4
-
very similar
4.5
-
5
identical
6.
Please rate the printed temporal bone for its value in surgical experience and training in comparison to temporal bone drilling in the operating room. 1
-
unlike
7.
1
poor
-
2.5
-
3
similar
-
3.5
-
4
-
very similar
4.5
-
5
identical
-
1.5
-
2
-
2.5
somewhat difficult
-
3
-
average
3.5
-
4
-
excellent
4.5
-
5
superior
-
1.5
-
2
-
few problems
2.5
-
3
-
average
3.5
-
4
-
excellent
4.5
-
5
superior
Please rate the printed temporal bone for its overall value as a surgical simulation, in preparation for cases in the OR. 1
poor
10.
2
Please rate the printed temporal bone for “safety” in the temporal bone laboratory in comparison to a cadaveric temporal bone. 1
9.
-
very different
Please rate the printed temporal bone for “ease of use” (setup in holder, preparing to drill, etc.) in the temporal bone laboratory in comparison to a cadaveric temporal bone. poor
8.
1.5
-
1.5
-
2
-
some value
2.5
-
3
-
average
3.5
-
4
-
excellent
4.5
-
5
superior
Did you experience any irritation (skin, eyes, lungs, mucous membranes) due to simulated bone dust in comparison to a cadaveric temporal bone? (assumes use of gown, gloves, mask and eye protection).
Yes
-
No
Figure 4
Figure 5
