The relationship between auditory efferent function and frequency selectivity in Man

The auditory efferent system consists of two populations of fibres. The lateral system connects with the inner hair cells and the medial system connects with the outer hair cells. Past work has suggested that the efferent system may act via the action of the outer hair cells to control cochlear mechanics and, in doing so, contribute to the high degree of frequency selectivity observed in the auditory system. This hypothesis has not, in the past, been fully examined in humans. In this study, normally hearing human subjects were examined to see if there was a link between efferent activity and frequency selectivity. The first experiment investigated whether there was a link between efferent action and frequency selectivity at 1kHz. Efferent activity was assessed using the contralateral suppression of transient and distortion product otoacoustic emissions (OAEs). Emissions were evoked by tones and clicks and suppressed by contralateral white noise, narrow band noise and tones. The suppression was examined in specific frequency bands as well as over the entire response range. Frequency selectivity was estimated using the notched-noise masking technique and 'roex' filter shape modelling. No conclusive relationship was found. The possibility that the efferent system might play a part in frequency selectivity only when activated by the contralateral ear was considered by testing the auditory filter during contralateral white noise. No consistent relationship was found between efferent function, and either the 1kHz filter shape measured during contralateral stimulation, or the change in filter shape. However, contralateral white noise did cause a significant broadening of the 1kHz filter via what was thought to be the action of the efferent system. The timing of the onset of the contralateral efferent effect was examined using OAEs in order to allow stimulus timings to be set so that the efferent system would be effective during filter shape testing. The onset latency was found to be between 17 and 20ms. Further analysis of the auditory filter shape during contralateral stimulation was undertaken by stimulating the contralateral ear with narrowband noises at 500Hz, 1kHz and 2kHz. This also provided no clear proof of a link with efferent activity. Finally, the filter shape at 2kHz was considered. Significant relationships emerged between the shape of the filter and the activity of the efferent system in this region. This implied that efferent control of frequency selectivity is not constant over the length of the basilar membrane. Overall, the effect of the efferent system on frequency selectivity in normally hearing humans was not found to be straightforward, and was found to vary with frequency. However, the results were consistent with the theory that the medial portion of the efferent system acts in such a manner as to dampen the motion of the basilar membrane.