Modeling rapid waveform compression on the basilar membrane as multiple-bandpass-nonlinearity filtering.

Evidence has accumulated from experimental intracochlear studies that nonlinear mechanical response of the basilar membrane is responsible for cochlear frequency tuning and is the major source of extracochlear nonlinear phenomena in cochlear sound analysis. Known basilar-membrane data provide a basis for synthesizing and quantifying conceptions of cochlear signal processing derived earlier from extracochlear studies that indicated the existence of rapid waveform compression and dual signal processing. The multiple-bandpass-nonlinearity (MBNL) model represents and generalizes available measurements of basilar-membrane mechanical responses in terms of a rapid nonlinear mixing at each place of an insensitive, linearlike lowpass filter with a sensitive, compressive bandpass filter. The dual filters are associated with the tails and tips of cochlear frequency tuning curves. Simulations of published nonlinear mechanical responses of the basilar membrane and predicted correlations with auditory-nerve responses are systematically explored. Correlations between model and biophysical data suggest that the model represents a nonlinear mixing by outer hair cells of hydromechanical and electromechanical signals, and thus provides a quantitative tool for biophysical study of cochlear mechanisms.