DPOAEs evoked by different stimulus paradigms in a fully nonlinear cochlear model

The DPOAE generation is a consequence of the intrinsically nonlinear nature of the cochlear dynamics. Different stimulus paradigms have been proposed for using DPOAE measurements as a diagnostic test of the hearing function. In particular, Kummer et al. [1] proposed a method for measuring the hearing threshold by extrapolating the growth rates of the linear DPOAE response at zero response magnitude, and defining this level as the best estimate of the audiometric threshold level. This method requires the so called “scissors” paradigm for evoking DPOAE. The scissors paradigm is considered capable of approximately maximizing the DPOAE response at any saturation regime, in all frequency ranges. In the scissors paradigm the L2 stimulus grows much faster than L1. A higher L1 level is generally needed because, in the “overlap” nonlinear distortion generation place, x(f2), the basilar membrane (BM) response to f2 is fully resonant whereas that to f1 is not. On the other hand, the nonlinearity of the BM response implies that the “advantage” of the f2 component decreases with increasing stimulus level, as the bandwidth of the response also increases. Different growth rates are obviously associated to DPOAE evoked by different paradigms. In addition, it is quite difficult to interpret the DPOAE growth curves when complex protocols are used to evoke them. In this work, a nonlinear non-local cochlear model is proposed to simulate otoacoustic emissions. The presence of strong nonlinearity, as a physical non-perturbative property of the system, requires a time domain solution of the equations representing the cochlea from a micromechanical point of view. The cochlear equations are solved in time domain by means of the state space variables mathematical formalism [2]. The model parameter space has been explored in order to generate DPOAE levels and growth rates compatible with experimental data on human subjects. The main properties of the DPOAE generated by the different paradigms have been reproduced. In particular, the “scissors” paradigm was able to maximize the DP amplitude at each L2 stimulus level, in agreement with the experimental evidence. The proposed model represents a useful tool for studying OAE evoked by complex protocols in strongly nonlinear regime