A modeling study of responses to sinusoidally amplitude-modulated stimuli in high frequency neurons in the auditory brainstem

Abstract

Human listeners take advantage of differences between the sounds arriving at the two ears to determine the location of sound sources. For high-frequency components, interaural time differences (ITDs) carried in the envelopes together with inter aural level differences (ILDs) are primary cues for sound localization. These binaural cues are initially extracted in the superior olivary complex (SOC) of the brainstem, which consists of two major nuclei: the medial superior olive (MSO) and the lateral superior olive (LSO). Olivary neurons, like all neurons in the ascending auditory pathway, project to the inferior colliculus (IC), where binaural information is integrated and reorganized before reaching higher nuclei for further processing. In the present study, physiologically explicit models of high-frequency neurons in the MSO, the LSO, and the IC were developed to investigate underlying mechanisms behind several empirical observations of neural responses in these nuclei. These models were applied to reproduce neural responses to sinusoidally amplitude-modulated (SAM) stimuli. Specifically, this thesis examined the following empirical observations: 1) many LSO neurons exhibit substantial decrease in discharge rate with increasing modulation frequency (fm) of monaural SAM tone stimuli; 2) some high-frequency IC neurons show dual types of envelope ITD sensitivity in response to binaural SAM tones, with peak-type responses for some modulation frequencies (fm) and trough-type responses for other frequencies; 3) some high-frequency IC neurons show dual types of envelope ITD sensitivity in response to binaural SAM noises. In this third case, the neurons exhibit one type of ITD sensitivity (either peak-type or trough-type) to cochlear-generated (CG) envelopes and the other type of ITD sensitivity to amplitude modulation (AM) envelopes. Results from the model developed in the present study show the following: 1) low threshold potassium channels, and in some cases background inhibition, may play the major role in the firing rate decrease with increasing modulation frequency in the LSO; 2) convergent inputs from the MSO and the LSO can account for the dual types of ITD sensitivity observed in some high-frequency IC neurons in response to binaural SAM tones, as long as the firing rates in the inputs from the MSO and the LSO are tuned to different fm; 3) the same neural connections (convergent inputs from MSO and LSO) can also explain the combination of peak-type and trough-type responses observed in the IC in response to SAM noises with correlated fine structures, in which case both CG and AM envelope ITDs are present. In addition to the simulation of empirical data, these models provide quantitative predictions which can be tested with future empirical experiments to further enrich our understanding of the neural basis of envelope processing.