Biophysical Population Models of the Auditory Nerve

Abstract

Biophysical models of the electrically stimulated auditory nerve are a powerful tool to simulate the neural response to cochlear implant stimulation. We review the history and development of these computational models, from single node Hodgkin-Huxley models to morphologically and physiologically inspired cable models of auditory nerve fibers. Next, we pair a stochastic, heterogeneous population model with two neurometric decision making paradigms to investigate the neural correlates of performance on the perceptual amplitude modulation detection task. The model predicts realistic modulation detection thresholds as a function of stimulus intensity and modulation frequency. When the decision criterion relates to fluctuations in instantaneous firing rate, low carrier rates are associated with better performance, but spectral analysis reveals that the modulation frequency is more strongly coded in the auditory nerve response at high carrier rates. Finally, we consider the effects of simulated pathology on the model’s population-scale response. We develop a single-fiber model of peripheral degeneration and compare temporal and stochastic properties of three distinct population models of cochlear pathology.