CInsertion - A Virtual Surgical Simulator for Training the Insertion of Intracochlear Electrodes

Purpose: Hearing health is one of the major variables that will determine quality of life and the fraction of the population with hearing conditions that could benefit from a Cochlear Implant (CI) is expected to grow considerably. Surgical training of the electrode insertion into the cochlea, the “blind” portion of the implantation surgery is currently limited, lacking integral feedback. It is key to prevent damage to the soft structures and detriment to the CI’s performance. To this end, CInsertion, a virtual reality, 3D, real time (RT) simulator for ENT specialist’s training is presented. Methods: Unity3D was used to develop the simulator alongside with 3D modeling programs for the Operating Room (OR) scene elements and morphological behaviour of the CI electrode. The cochlear model was based on high resolution segmented images of Open Ear Data Base and the temporal bone involved was CT scanned and processed in 3DSlicer. 3D Systems Touch Haptic Device was implemented as input for the spatial control of the electrode and the simulator’s final version runs on Windows. Results and user remote data bases were structured on MongoDB. Results: A functional simulator which aims at the training of the ability to become spatially aware and enhance the motion sensitivity of the trainee during CI electrode insertion was implemented. In a RT 3D pressure color map, the effect of friction against the cochlear structures is visualized and progress over time can be assessed. Qualitatively, the electrode´s bending shows good correspondence with the behaviour of surgical cases as compared with imaging. The legitimacy and applicability of the development was validated by a group of experienced surgeons. Conclusion: Every innovative component in this field will likely improve the final performance of the CI and so, of hearing, building an exciting path ahead. Future work involves quantitative validation, inclusion of non generic electrodes and patient specific model generation.