A computational algorithm for computing nonlinear auditory frequency selectivity

Computational algorithms that mimic the response of the basilar membrane mu se be capable of reproducing a range of complex features that are character istic of the animal observations. These include complex input output functi ons that are nonlinear near the site's best frequency, but linear elsewhere . This nonlinearity is critical when using the output of the algorithm as t he input to models of inner hair cell function and subsequent auditory-nerv e models of low- and high-spontaneous rate fibers. We present an algorithm that uses two processing units operating in parallel: one Linear and the ot her compressively nonlinear. The output from the algorithm is the sum of th e outputs of the linear and nonlinear processing units. Input to the algori thm is stapes motion and output represents basilar membrane motion. The alg orithm is evaluated against published chinchilla and guinea pig observation s of basilar membrane and Reissner's membrane motion made using laser veloc imetry. The algorithm simulates both quantitatively and qualitatively, diff erences in input/output functions among three different sites along the coc hlear partition. It also simulates quantitatively and qualitatively a range of phenomena including isovelocity functions, phase response, two-tone sup pression, impulse response, and distortion products. The algorithm is poten tially suitable for development as a bank of filters, for use in more compr ehensive models of the peripheral auditory system. (C) 2001 Acoustical Soci ety of America.