Electrocochleography Predictions Using a Combined Model of Acoustic Hearing and Electric Current Spread in the Cochlea

Background: Electrocochleography (eCochG) through the implant is advocated as a tool to monitor the residual hearing during cochlear implant surgery. Postoperatively, this eCochG tool can objectively asses hearing thresholds in selected parts of the cochlea by recording responses to pure tones.

Objective: At present it is not known exactly how hair cell and neural activity contribute to eCochG responses detected from different locations in the cochlea. In the current study a model is evaluated that simulates CM responses to various stimuli at different intra-cochlear locations.

Methods: Electric field spread of electric dipoles, representing inner and outer hair cells, at 3200 spatially different locations on the basilar membrane was calculated using a 3D volume conduction model (Kalkman et al., 2015). The hair cell activation was calculated using an online available phenomenological model of the complete auditory periphery (Bruce et al., 2018). Both models were combined to predict the cochlear microphonic (CM).

Results: The predicted local hair cell’s electric field closely resembled the exponential decay, inferred from intra-operative recordings. CM responses over the array show a narrowing peak that shifts basally with increasing stimulus frequency. The fact that CM responses to low-frequency tones can be recorded along the whole array is a result of cross-turn sensitivity rather than from excitation of basal hair cells.

Conclusions: These simulations show that the intracochlearly recorded eCochG is a measure of very local activation, while cross-turn sensitivity broadens eCochG activation for lower frequencies. Therefore, ECochG is applicable to measure acoustic tuning curves of the implanted cochlea.