Simulation of auditory–neural transduction: Further studies

Abstract

A computational model of mechanical to neural transduction at the hair cell-auditory-nerve synapse is presented. It produces a stream of events (spikes) that are precisely located in time in response to an arbitrary stimulus and is intended for use as an input to automatic speech recognition systems as well as a contribution to the theory of the origin of auditory-nerve spike activity. The behavior of the model is compared to data from animal studies in the following tests: (a) rate-intensity functions for adapted and unadapted responding; (b) two-component short-term adaptation; (c) frequency-limited phase locking of events; (d) additivity of responding following stimulus-intensity increases and decreases; (e) recovery of spontaneous activity following stimulus offset; and (f) recovery of ability to respond to a second stimulus following offset of a first stimulus. The behavior of the model compares well with empirical data but discrepancies in tests (d) and (f) point to the need for further development. Additional functions that have been successfully simulated in previous tests include realistic interspike-interval histograms for silence and intense sinusoidal stimuli, realistic poststimulus period histograms at various intensities and nonmonotonic functions relating incremental and decremental responses to background stimulus intensity. The model is computationally convenient and well suited to use in automatic recognition devices that use models of the peripheral auditory system as input devices. It is particularly well suited to devices that require stimulus phase information to be preserved at low frequencies.

A phenomenological model of the synapse between the inner hair cell and auditory nerve: Long-term adaptation with power-law dynamics

“…Although the mechanism that gives rise to synaptic adaptation is not completely understood, it could be caused either by the depletion of neurotransmitter from a readily releasable presynaptic pool of neurotransmitter ͑Moser and Beutner, 2000; Schnee et al, 2005;Goutman and Glowatzki, 2007͒ or by the desensitization of post-synaptic receptors ͑Raman et al, 1994͒. Modeling the adaptation in the IHC-AN synapse has been a focus of extensive research over the last several decades. Early attempts employed a single-reservoir system with loss and replenishment of transmitter quanta ͑Schroeder and Hall, 1974;Sujaku, 1974, 1975͒, and later models added extra reservoirs ͑or sites͒ or more complex principles of transmitter flow control ͑Furukawa and Matsuura, 1978;Furukawa et al, 1982;Ross, 1982Ross, , 1996Schwid and Geisler, 1982;Smith and Brachman, 1982;Cooke, 1986;Meddis, 1986Meddis, , 1988Westerman and Smith, 1988͒. In general, the transmitter in these models lies in reservoirs or sites close to the presynaptic membrane and diffuses between reservoirs within the cell and out of the cell to the synaptic cleft.…”

“…The diversity and complexity of adaptation pose a great challenge for successful modeling of the dynamics of this synapse. Two models with different structures have been developed independently in a series of studies ͑Meddis, 1986͑Meddis, , 1988Westerman and Smith, 1988;Carney, 1993;Zhang et al, 2001;Sumner et al, 2002. However, the mathematical descriptions of these two models are essentially equivalent despite their structural differences ͑Zhang and Carney, 2005͒.…”

A Functional Point-Neuron Model Simulating Cochlear Nucleus Ideal Onset Responses

“…The output of each gammatone filter serves as input to Meddis' model of the mechanical to neural transduction at the hair cell-auditory nerve synapse (Meddis, 1986(Meddis, , 1988Meddis et al, 1990). The output is the instantaneous AN discharge rate function in response to arbitrary stimuli.…”

“…In the peripheral stage of the present model, the mechanical to neural transduction at the hair cell-auditory nerve synapse was simulated using Meddis' model implementation (Meddis, 1986(Meddis, , 1988Meddis et al, 1990). Instead of actually computing individual spike trains for each AN fiber the "deterministic" rate function at the output of the synapse was considered as representing the AN activity.…”

Separation of speech from interfering sounds based on oscillatory correlation

“…In the final stage of the peripheral model, the output of each gammatone filter is processed by the Meddis [32] model of inner hair cell function. The output of the hair cell model is a probabilistic representation of firing activity in the auditory nerve, which incorporates well-known phenomena such as saturation, two-component short-term adaptation and frequency-limited phase locking.…”