Grant: Collaborative Health Research Projects

Mastoidectomy is a complex surgical procedure in which portions of the skull are drilled awayto treat ear infections, to place cochlear implants and to remove tumours. Hundreds ofthousands of mastoidectomies are performed worldwide each year with surgical complicationrates as high as 20%. Surgical mistakes are disastrous and include facial nerve paralysis,hearing loss, tinnitus, vertigo and altered taste. Significant training is required to master theprocedure. However, training is hampered by a shortage of cadaveric materials. To providean alternative training approach, we have developed a virtual-reality simulator that allows theuser to operate on a single 3D digital head model. Practice on a single model does notprepare trainees for the large variability in patient anatomy, but creating additional models isdifficult. Moreover, careful evaluation of the simulator is required for acceptance.Our objectives are to (1) develop software to automatically construct digital models frompatient images, (2) implement automated performance assessment methods, and (3)evaluate skill development and transfer from the virtual environment to realistic settings.We will develop software to automatically warp a template digital anatomical model to matcha specific patient image, thus generating models for new patients. We will implementsoftware that automates evaluation of each trainees performance when operating on virtualhead models. The software will be based on well validated metrics that are used in currentpractice by instructors. To evaluate skill development and transfer, an expert panel ofOtolaryngologists will evaluate trainees on cadaveric skulls before and after training using thesimulator.The simulator will result in increased and improved training opportunities on a wide variety ofpatient types. Moreover, it will allow rehearsal on a digital model prior to surgery on a specificanatomy, leading to improved surgical performance.