An Improved Stochastic Hodgkin-Huxley Based Model of a Node of Ranvier for Cochlear Implant Stimulation

Abstract

Cochlear implants (CIs) are prosthetic devices used to partially restore hearing for profound and severely deaf individuals. CIs convert sounds into electrical pulses which stimulate the auditory nerve fibers. An accurate model of auditory nerve fibers (ANFs) would help in improving the functionality of CIs. Previous studies have shown that the original Hodgkin-Huxley (1952) model (with kinetics adjusted for mammalian body temperature) may be better at describing nodes of Ranvier in ANFs than models for other mammalian axon types. However, the Hodgkin-Huxley model is still unable to explain a number of phenomena observed in auditory nerve responses to CI stimulation, such as short-term and long-term adaptation, the time-course of relative refractoriness, and stimulus-dependent random fluctuations in membrane threshold. Recent physiological investigations of spiral ganglion cells have shown the presence of a number of ion channel types not considered in the previous modelling studies, including low-threshold potassium (𝐼^KLT) channels and hyperpolarization-activated cation (𝐼^h) channels. In this thesis, inclusion of these ion channel types in a stochastic Hodgkin-Huxley model is investigated. Four versions of the model are formed and compared: that is, the standard Hodgkin-Huxley model, the standard model with /h only added, the standard model with 𝐼^KLT only added, and finally, the standard model with both h and 𝐼^KLT added. Two group of responses are explored: i) single-pulse responses and ii) pules-train responses. For the single pulse responses, a charge-balanced biphasic stimulus pulse is used. The effect of varying the pulse-width and the interphase gap is investigated for both leading phase polarities. Results are compared to responses for single monophasic stimulus pulses in some cases. Pulse-train responses are investigated for charge-balanced depolarizing-phase leading biphasic pulses at rates of 200, 800, and 2000 pulse/s. Results from single-pulse responses show an increase in spike threshold when one or both of these channel types are included. The addition of 𝐼^KLT increases random threshold fluctuations in the stochastic model, particularly for longer pulse widths. For pulse-train responses, rapid adaptation in spike rate may be resulting from 𝐼^KLT whereas 𝐼^h produces slower "short-term" adaptation. Thus, the simulation results suggest that including 𝐼^KLT and/or 𝐼^h in a Hodgkin-Huxley model improves the accuracy of the model in describing auditory nerve fiber responses during cochlear implant stimulation.