Comparison of cadaveric and isomorphic virtual haptic simulation in temporal bone training
Journal of Otolaryngology - Head & Neck Surgery volume 43, Article number: 31 (2014)
Virtual surgery may improve learning and provides an opportunity for pre-operative surgical rehearsal. We describe a novel haptic temporal bone simulator specifically developed for multicore processing and improved visual realism. A position locking algorithm for enhanced drill-bone interaction and haptic fidelity is further employed. The simulation construct is evaluated against cadaveric education.
A voxel-based simulator was designed for multicore architecture employing Marching Cubes and Laplacian smoothing to perform real-time haptic and graphic rendering of virtual bone.
Ten Otolaryngology trainees dissected a cadaveric temporal bone (CTB) followed by a virtual isomorphic haptic model (VM) based on derivative microCT data. Participants rated 1) physical characteristics, 2) specific anatomic constructs, 3) usefulness in skill development and 4) perceived educational value. The survey instrument employed a Likert scale (1-7).
Residents were equivocal about the physical properties of the VM, as cortical (3.2 ± 2.0) and trabecular (2.8 ± 1.6) bone drilling character was appraised as dissimilar to CTB. Overall similarity to cadaveric training was moderate (3.5 ± 1.8). Residents generally felt the VM was beneficial in skill development, rating it highest for translabyrinthine skull-base approaches (5.2 ± 1.3). The VM was considered an effective (5.4 ± 1.5) and accurate (5.7 ± 1.4) training tool which should be integrated into resident education (5.5 ± 1.4). The VM was thought to improve performance (5.3 ± 1.8) and confidence (5.3 ± 1.9) and was highly rated for anatomic learning (6.1 ± 1.9).
Study participants found the VM to be a beneficial and effective platform for learning temporal bone anatomy and surgical techniques. They identify some concern with limited physical realism likely owing to the haptic device interface. This study is the first to compare isomorphic simulation in education. This significantly removes possible confounding features as the haptic simulation was based on derivative imaging.
Current temporal bone surgical training is centered on graduated operative practice under the supervision of an experienced surgeon. As a corollary to increasing focus on safety, and to supplement surgical education in the face of resident work hour restrictions, numerous teaching adjuncts have been developed. The Cadaveric Temporal Bone Lab remains the gold standard; however access to sufficient exposure is site specific owing to local factors and expense . An array of haptic simulators - are now available to complement this training and the field of additive manufacturing is beginning to provide effective models for dissection ,.
Haptic simulation for surgical training
Haptic simulation provides real-time 3 dimensional contact force representation. The user sees a graphical representation of the bone and feels it using a manipulandum held in the hand in an analogous fashion to an otic drill. Movement of the manipulandum guides the virtual drill tip. As the virtual bone is drilled, deep structures are revealed, permitting simulated complex surgical procedures. While this does provide a sense of drill-bone interaction, the experience is not identical to that of operative drilling.
The advantages of haptic simulation are easy operation, the absence of biologic materials, the ability to provide a wide range of anatomic variants, failure without consequence, and provision for repeated practice. Perhaps the most significant advantage is the ability to objectively monitor and assess trainee actions, providing a basis for formative and summative metrics -. Further, there may be utility in competency based residency training.
These benefits have led to the development of numerous haptic surgical trainers -. The validity of haptic trainers has been studied, particularly with reference to surgical performance and construct validity ,-. Direct comparison to performance in standard cadaveric dissection , have previously shown mixed results. These studies appear to have used a standardized haptic model which was compared to anatomically unmatched cadaveric samples. In the study described below, unique isomorphic models of cadaveric bones were created so that participants drill anatomically identical bones in both modalities, eliminating anatomical variation as a confounding factor in analysis.
The haptic simulation of temporal bone which we use takes advantage of incremental gains in processing speed and computer architecture to generate contact forces using a novel algorithm .
Haptic simulation of the temporal bone
Temporal bone haptic simulation is not new. The earliest simulators converted voxel data to low resolution polygon surfaces  for display using volume rendering -. Our current simulation also uses voxel data for collision detection and force calculations, but renders the voxels graphically using high resolution polygons generated by the Marching Cubes algorithm  and Laplacian HC Smoothing . These two algorithms run in real time using a new multicore architecture, creating a bone surface which appears smooth and free of step-like voxellation artifacts (Figure 1). The simulation runs on the Windows’ platform using DirectX ™ which allows stereoscopic 3D using inexpensive consumer level 3D graphics cards and active shutter displays.
Our haptic display simulates forces felt by a simulated surgical drill. For the purposes of this study we used the inexpensive Phantom Omni device (Geomagic, Wilmington MA). The program is also compatible with 6 degree-of-freedom devices.
A position locking algorithm is used to calculate interaction forces rather than the more commonly used virtual spring methods . This permits calculation of the location of the drill bit at every iteration and allows the haptic device to navigate fine surface features and improve stability when the drill tip is located in tightly constrained spaces.
The temporal bone haptic simulation we have developed employs CT data. The data is segmented into component structures, stored initially as individual polygon meshes which are then combined into a voxellated model for haptic display.
Education centric platform
The purpose of the haptic simulation is to aid education. Software features included in the simulation permit drilling actions to be undone at the discretion of the user. Internal constructs can be made “undrillable” to facilitate learning the relative nature of anatomy. The ease of bone removal can be modified to aid in learning structure location. Two distinct training modes permit a user to both visually and manually follow an expert’s dissection of a bone model. The first is Passive Hand Motion Training. In this mode the user holds the haptic manipulandum while the computer replays the exact drill movements of an expert. The second mode is Active Hand Motion Training where arrows located in the upper right of the screen direct hand motion to closely replicate the expert’s recorded drilling process. Variable coloration, transparency, and stiffness of individual tissue components permit users to visualize anatomic structures more easily (Figures 2 and 3).
In the following, we describe the design of derivative haptic simulations from cadaveric temporal bone microCT data followed by experimental design, surgical resident preferences and perceptions of the model when compared directly to matched isomorphic cadaveric specimens.
After study approval by the local Research Ethics Board (REB), ten residents each performed a cadaveric immediately followed by a virtual dissection of an isomorphic haptic model.
Preparation of isomorphic haptic models from cadaveric specimens
Ten human cadaveric temporal bone specimens were prepared for otic drilling by resident surgical trainees. Prior to drilling, each bone underwent microCT using a SkyScan 1176 microtomograph (Bruker-microCT, Belgium). Image resolution was initially 35 μm but was down-sampled by a factor of 4 in x and y.
MicroCT data was then segmented using Mimics 22.214.171.124 (Materalize, Belgium) into separate anatomic features. Bone was segmented semi-automatically using Hounsfield unit thresholds. This ensured that void spaces such as air cells, were retained in the final model. Soft tissue features including carotid artery, sigmoid sinus, superior petrosal sinus, dural plates, endolyphatic sac, endolymphatic duct, otic capsule contents, ossicles, greater superficial petrosal, chordae tympani, facial nerves, cochleariform process and semi-canal for tensor tympani, were manually segmented. Segmented features were stored as individual polygon meshes.
A haptic simulation of each cadaveric specimen was then generated by recombining its individual polygon mesh models into a single voxellated model . Each cadaveric bone specimen, therefore, had a corresponding haptic simulation which contained anatomy identical in size and shape (isomorphic) to the parent bone. The simulation used a haptic device (Geomagic Touch - SC, USA) to control a virtual drill during interaction with the voxellated model (Figures 1, 2 and 3). The model was visually displayed on a 165 cm plasma screen (Panasonic TCP65VT30, Panasonic, Osaka, Japan) mounted above and behind the haptic device (1280 × 720 pixel resolution). The drill was activated using an on-off foot-pedal (Scythe – Tokyo, Japan).
Resident evaluation of haptic temporal bone models
10 surgical resident trainees, with varying degrees of surgical experience, from the Otolaryngology program at the University of Manitoba, gave informed consent for participation in the study. Each student was randomly assigned a cadaveric bone and its matched isomorphic haptic model for dissection. Subjects first drilled their assigned cadaveric specimen under supervision of a Neurotologist using an otic drill (Stryker, Michigan, USA). Following completion of cadaveric drilling, each subject drilled the isomorphic haptic model matching the cadaveric bone on which they had just practiced. No time limit on the session was set although all subjects completed cadaveric and virtual drilling in less than 4 hours. Subjects then completed a survey instrument (Likert Scale) comparing haptic and cadaveric drill experiences. The survey asked subjects to rate the haptic model in four areas as compared to cadaveric dissection, including 1) physical characteristics of the VM, 2) specific anatomic feature representation of the VM 3) usefulness in surgical skills training and 4) perceived educational value. A copy of the survey instrument can be obtained from the corresponding author.
The mean and standard deviation of resident responses can be seen in the tables below for each of the four survey components.
Residents were ambivalent about the physical similarity of the VM as compared to CTB (Table 1), rating it highest for its air cell system representation (5.4 ± 1.4). Hardness was rated better for cortical (3.2 ± 2.0) than trabecular (2.8 ± 1.6) bone, but neither was considered similar to CTB. The simulations overall physical similarity to CTB (3.5 ± 1.8) was unexceptional.
Residents generally rated the VM’s internal constructs as more similar to CTB than its physical properties (Table 2), with the highest values awarded to vascular structures (range 5.6 to 5.8) and the lowest to dural plates (4.5 ± 1.7). Important middle ear, otic capsule and nervous structures were reasonably considered (range 5-5.5).
Residents generally felt that the VM was beneficial in surgical skill acquisition (Table 3), rating it highest for translabyrinthine approaches to the skull base (5.2 ± 1.3) and lowest for sigmoid sinus decompression (4.4 ± 2.0). All surgical skills assessed were deemed to benefit from training on the VM.
Residents generally agreed that the VM was an effective (5.4 ± 1.5) and accurate (5.7 ± 1.4) tool which should be integrated into education (5.5 ± 1.4) (Table 4). Participants did not consider the VM a viable replacement of CTB dissection (2.5 ± 2.3). Generally, the VM was presumed to increase surgical performance (5.3 ± 1.8) and confidence (5.3 ± 1.9) and was ranked highly with respect to its usefulness in teaching anatomy (6.1 ± 1.9) and facilitating access to a broad range of pathologic and anatomic variation (5.6 ± 1.8).
This is a first description of a novel multicore haptic temporal bone simulation employing a position locking algorithm and validated using isomorphic models. The simulation allows for multiple segmented models to be created in the formation of a virtual library.
Participants feel the haptic simulation is beneficial in learning surgical skills and neurotologic surgical approaches. Residents found the haptic simulation to be an effective teaching platform with favorable internal anatomic representation.
Problematic drill character is due to the inability of the haptic device to render a stiff bone surface with realistic drill vibration. While improving digital processing and graphic representation, a multicore design has limited impact on the effectiveness of the manipulandum. The change to a position locking algorithm, realizes improved processing; however, based on these results, does not further advance drill experience. A direct comparison to a virtual spring haptic system was not undertaken, but may prove useful to determine differences in the user experience.
This study has several advantages. The most noteworthy is the use of an isomorphic haptic model, derived directly from the template CTB. Previous studies have focused on comparing generic cadaveric dissection to haptic simulation ,. Cadaveric dissection irreversibly destroys the anatomy of the specimen. By preserving the specimen’s anatomy in the form of a haptic model, it is possible to make direct comparison of the simulation’s effectiveness without the confounding issue of differing anatomical features between the studied modalities. The preservation of the anatomy also permits repeated testing on the same specimen and the potential for developing large digital libraries.
The most serious study limitation is its small sample size and single centre nature. While the study examined all ENT surgery residents at the University of Manitoba, it is possible that institutional bias may have influenced the findings. A more rigorous multi-centre trial design with a carefully defined curricular program, looking at resident and expert perceptions, as well as performance metrics and clinical outcomes is currently being designed.
Improved training may reduce patient risk exposure. Simulated temporal bone training can address needs in continuing education, competency based residency training, and ultimately become a component of the certification process. We currently have a large and expanding library based on microCT data which we hope to utilize for these purposes.
We describe a novel haptic temporal bone simulator (VM) derived from imaging of cadaveric bone. We evaluated our VM against the derived cadaveric bone. Study participants found that the VM was both a beneficial and an effective platform for learning temporal bone anatomy and surgical techniques. They also identify some concern with limited physical realism likely owing to the haptic device interface. Virtual surgery may improve learning and provide added opportunity for pre-operative surgical rehearsal without comparison patient safety. This study is the first to compare isomorphic simulation in education.
DW was involved with study design, conducting the study and drafting the paper. BU was involved with study design and statistical analysis. JK designed the virtual haptic model. JP was involved with statistical analysis. CR was involved with study design. JBH was involved with study design, drafting the paper and supervising the study team. All authors read and approved the final manuscript.
Cadaveric temporal bone
Virtual isomorphic haptic model
Fennessy BG, Sullivan PO: Establishing a temporal bone laboratory: considerations for ENT specialist training. Ir J Med Sci. 2009, 178: 393-395. 10.1007/s11845-009-0373-x.
Fried MP, Uribe JI, Sadoughi B: The role of virtual reality in surgical training in otorhinolaryngology. Curr Opin Otolaryngol Head Neck Surg. 2007, 15: 163-169. 10.1097/MOO.0b013e32814b0802.
George AP, De R: Review of temporal bone dissection teaching: how it was, is and will be. J Laryngol Otol. 2010, 124: 119-125. 10.1017/S0022215109991617.
Morris D, Sewell C, Barbgli F, Salisbury K, Blevins NH, Girod S: Visuohaptic simulation of bone surgery for training and evaluation. IEEE Comput Graph Appl. 2006, 6: 48-57. 10.1109/MCG.2006.140.
Wiet GJ, Bryan J, Dodson E, Sessanna D, Stredney D, Schmalbrock P, Welling B: Virtual temporal bone dissection simulation. Stud Health Tech Informat. 2000, 70: 378-384.
Pflesser B, Petersik A, Tiede U, Höhne KH, Leuwer R: Volume cutting for virtual petrous bone surgery. Comput Aided Surg. 2002, 7 (2): 74-83. 10.3109/10929080209146018.
Agus M, Giachetti A, Gobbetti E, Zanetti G, Zorcolo A, John NW, Stone RJ: Mastoidectomy simulation with combined visual and haptic feedback. MMVR. 2002, 85: 17-23.
Sorensen MS, Mosegaard J, Trier P: The visible ear simulator: a public PC application for GPU-accelerated haptic 3D simulation of ear surgery based on the visible ear data. Otol Neurotol. 2009, 30 (4): 484-487. 10.1097/MAO.0b013e3181a5299b.
Jackson A, John NW, Thacker NA, Ramsden RT, Gillespie JE, Gobbetti E, Zanetti G, Stone R, Linney AD, Alusi GH, Franceshini SS, Scherdtner A, Emmen A: Developing a virtual reality environment in petrous bone surgery: a state-of-the-art review. Otol Neurotol. 2002, 23: 111-121. 10.1097/00129492-200203000-00001.
Zhao YC, Kennedy G, Yukawa K, Pyman B, O’Leary S: Improving temporal bone dissection using self-directed virtual reality simulation: results of a randomized blinded control trial. Otolaryngol Head Neck Surg. 2011, 144 (3): 357-364. 10.1177/0194599810391624.
Hochman J, Kraut J, Kazmerick K, Unger B: Generation of 3D printed temporal bone model with internal fidelity and validation of the mechanical construct. Otolaryngol Head Neck Surg. 2013, 150 (3): 448-454. 10.1177/0194599813518008.
Roosli C, Sim JH, Mochel H, Mokosch M, Probst R: An artificial temporal bone as a training tool for cochlear implantation. Otol Neurotol. 2013, 34 (6): 1048-1051. 10.1097/MAO.0b013e31828f4907.
Zirkle M, Roberson DW, Leuwer R, Dubrowski A: Using a virtual reality temporal bone simulator to assess otolaryngology trainees. Laryngoscope. 2007, 117: 258-263. 10.1097/01.mlg.0000248246.09498.b4.
Reddy-Kolanu G, Alderson D: Evaluating the effectiveness of the voxel-man TempoSurg virtual reality simulator in facilitating learning mastoid surgery. Ann Roy Coll Surg Engl. 2011, 93: 205-208. 10.1308/003588411X565987.
Wiet GJ, Stredney D, Kerwin T, Hittle B, Fenandez SA, Abdel-Rasoul M, Welling B: Virtual temporal bone dissection system: development and testing. Laryngoscope. 2012, 122 (Suppl 1): S1-S12. 10.1002/lary.22499.
Francis HW, Malik MU, Varela DADV, Barffour MW, Chien WW, Carey JP, Niparko JK, Bhatti NI: Technical skills improve after practice on virtual-reality temporal bone simulator. Laryngoscope. 2012, 122: 1385-1391. 10.1002/lary.22378.
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: 1225-1231. 10.1097/MAO.0b013e31825e7977.
Kerwin T, Wiet G, Stredney D, Shen H: Automatic scoring of virtual mastoidectomies using expert scoring. Int J Comput Assist Robot Surg. 2012, 7: 1-11. 10.1007/s11548-011-0566-4.
Kraut J, Hochman JB, Unger B: Temporal bone surgical simulation employing a MultiCore architecture. In Proceedings of the 26thAnnual IEEE Canadian Conference on Electrical and Computer Engineering: 5 - 8 May 2013; Regina, SK. Edited by Yao J, Mehrandezh M, Paranjape R, Gelowitz G. IEEE; 2013:1-6.
Morris D, Sewell C, Blevins N, Barbagli F, Salisbury K: A collaborative virtual environment for simulation of temporal bone surgery. MICCAI. 2004, 3217: 319-327.
Bryan JA: A Virtual Temporal Bone Dissection Simulation. Master of Science Thesis. 2001, Ohio State University, Department of Computer and Information Science, Columbus, Ohio
Visible Ear Simulation Manual. [ November 2012], [http://www.daimi.au.dk/~trier/VES_blog/]
Leuwer R, Petersik A, Pflesser B, Pommert A, Tolsdorff B, Hohne KH, Tiede U: Voxel-man TempoSurg: a virtual reality temporal bone surgery simulator. J Jpn Soc Head Neck Surg. 2007, 17 (3): 203-207. 10.5106/jjshns.17.203.
Lorensen WE, Cline HE: Marching cubes: a high resolution 3D surface construction algorithm. Comput Graph. 1987, 21 (4): 163-169. 10.1145/37402.37422.
Vollmer J, Mencel R, Muller H: Improved laplacian smoothing of noisy surface meshes. Comput Graph Forum. 1999, 18 (3): 131-138. 10.1111/1467-8659.00334.
Research Support: Stryker donated an Otic Drill System. No Industrial Support for Research. No Disclosures to report. No Conflict of Interest.
About this article
Cite this article
Wong, D., Unger, B., Kraut, J. et al. Comparison of cadaveric and isomorphic virtual haptic simulation in temporal bone training. J of Otolaryngol - Head & Neck Surg 43, 31 (2014). https://doi.org/10.1186/s40463-014-0031-9