A Combined Model for Electrocochleography Simulations in Cochlear Implant Users

Authors: Aristeidis Choustoulakis1

1Leiden University Medical Center


Background: Electrocochleography (ECochG) recordings have gained a lot of attention as an assessment tool to reduce surgical trauma during cochlear implantation surgery. A combined model has been proposed to include factors for the purpose of ECochG simulations in CI users. To validate the performance of the model, it is compared against the previously used model and animal experimental data. The effect of different biophysical factors of the auditory periphery on the various model outputs is explored.

Methods: The proposed computational model combines different stages. The middle ear filter and the basilar membrane are modelled using filters whereas the inner hair cells and the outer hair cells are modelled biophysically, via their circuit representation. The auditory nerve fiber and the ECochG stages are modelled mathematically. The outputs compared include: input-output curves of the middle ear to the basilar membrane, frequency responses of the hair cells at different sound levels, saturation behavior of the hair cells, auditory nerve fiber responses and tuning curves, and cochlear microphonics simulations.

Results: Outputs showed similarities to experimental animal data and better followed the physiological behavior compared to the previous model. The proposed model better simulates the saturation behavior of the hair cells compared to the previous model which did not include saturation. The frequency response of the hair cells in the new model show similar levels compared to the experimental data whereas the levels in the previous model are one order of magnitude higher. Electrode frequency mapping was more accurate in the proposed model.

Conclusion: The proposed model that was developed follows physiology more closely compared to the previous phenomenological model. When comparing outputs of the models to the experimental data, the proposed model matches the level of the experiments more closely compared to the previous phenomenological model.