Extraction of spectral information in auditory brain stem nuclei; hypothesis and experimental observations

Abstract

Psychophysical evidence strongly suggests that there is a central neural pattern of representation of sound frequencies; and that the overall pitch of any sound is a function of this central pattern. Physiological studies on the discharges of auditory nerve fibers as a function of cochlear location, and psychophysical studies employing controlled electrical stimulation of the auditory nerve array in man strongly indicate that the spectral composition of complex sounds is not extracted from either discharge rate versus position profiles across the auditory nerve array (the “cochlear place”), or by measuring input discharge periodicities per se. It is hypothesized that spectral components of sounds are extracted from the relative timing of discharges of auditory nerve fibers from separated locations along the auditory nerve array. To extract a spectral representational pattern with broad dynamic range, comparator neurons must detect the synchronicity of arrival of input from separated basilar membrane locations with a resolution of 10–20 μs. It is hypothesized that this extraction is effected by the neurons of the medial superior olive, which operate as monaural as well as binaural time comparators. Psychophysical and physiological data supporting testable predictions of this hypothesis will be reviewed. [Work supported by NIH Contract N01-NS-7-2367 and NIH Grant NS-11804.]