A Phenomenological Model Reproducing Temporal Response Characteristics of an Electrically Stimulated Auditory Nerve Fiber

Abstract

The ability of cochlear implants (CIs) to restore hearing to profoundly deaf people is based on direct electrical stimulation of the auditory nerve fibers (ANFs). Still, CI users do not achieve as good hearing outcomes as their normal-hearing peers. The development and optimization of CI stimulation strategies to reduce that gap could benefit from computational models that can predict responses evoked by different stimulation patterns, particularly temporal responses for coding of temporal fine structure information. To that end, we present the sequential biphasic leaky integrate-and-fire (S-BLIF) model for the ANF response to various pulse shapes and temporal sequences. The phenomenological S-BLIF model is adapted from the earlier BLIF model that can reproduce neurophysiological single-fiber cat ANF data from single-pulse stimulations. It was extended with elements that simulate refractoriness, facilitation, accommodation and long-term adaptation by affecting the threshold value of the model momentarily after supra- and subthreshold stimulation. Evaluation of the model demonstrated that it can reproduce neurophysiological data from single neuron recordings involving temporal phenomena related to inter-pulse interactions. Specifically, data for refractoriness, facilitation, accommodation and spike-rate adaptation can be reproduced. In addition, the model can account for effects of pulse rate on the synchrony between the pulsatile input and the spike-train output. Consequently, the model offers a versatile tool for testing new coding strategies for, e.g., temporal fine structure using pseudo-monophasic pulses, and for estimating the status of the electrode-neuron interface in the CI user's cochlea.

Models of Cochlea Used in Cochlear Implant Research: A Review

“…In contrast to the biophysical approach, phenomenological models do not rely on specific biophysical mechanisms and derive empirical relationships based on neurophysiological and psychophysical observations [141]. Due to the much-reduced parameter space, this approach allows the efficient modelling of complex phenomena that can be adjusted to individual CI patients and has proven effective at predicting and explaining a diverse range of auditory phenomena [139,141,142].…”

A model framework for simulating spatial hearing of bilateral cochlear implant users

“…The present model is insensitive to the number and polarity of pulse phases, and to the size of the inter-phase gap. Several studies have investigated the polarity sensitivity of auditory nerve fibers using eCAP recordings in human CI users [112,113], and to model such experiments, other AN models (e.g., [19,20,114]) can be incorporated. (3) Extensions to predict bilateral speech intelligibility are possible by connecting the AN spike outputs to the backend processing stages of other purposes such as an automatic speech recognition system (e.g., [37,110,111]) or by replacing it with more abstract internal model representations (e.g., [115,116]).…”