Model of the Encoding Function of Primary Auditory Neurons

Abstract

A model of the encoding function of primary auditory neurons has evolved from analysis of electrophysiological data. The model assumes that a primary neuron is exposed to cyclical inflections of a small length of the basilar membrane and that the neuron possesses a thresholding property that interacts with these inflections. It is also assumed that the neuron responds to inflections in one direction only and, therefore, possesses some degree of phase locking. The threshold of the neuron is not time-constant, but has the general shape of a decaying exponential of the form e−kt (t=0 indicates time of last firing of this neuron). The value of k is “noisy,” fluctuating from one discharge to the next, but maintains some average value. The average value of k determines whether the neuron encodes frequency, amplitude, or a combination of both. The average value of k is influenced by the immediately preceding average-pulse-rate output of the neuron, which gives to the neuron some of the properties of a feedback system. Equally important is the concept that a single neuron must function as a part of a population and that accurate retrieval of information is accomplished by spatial averaging of the population. Response of the model to auditory signals is quantitatively compared to electrophysiological data.