A nonlinear mechano-electro-acoustic model of the human cochlea

While direct measurements of the basilar membrane (BM) and auditory nerve fiber tuning are possible in lab animals [1] but no psychoacoustic measurements (with rare exceptions [2]), only otoacoustic emission and psychophysical tuning curve measurements [2,3,4] are possible in humans. In an attempt to partly address this missing link, the three-dimensional physiology-based nonlinear mechanical-electrical-acoustic (MEA) model of the human cochlea is developed in this study to enable prediction of both otoacoustic emissions as well as BM and organ of Corti vibration tuning. The nonlinear MEA model of the human cochlea adapts our prior linear MEA model of the guinea pig cochlea [5,6] to humans and includes the nonlinearity due to mechano-electrical transduction in hair bundles. The transmission line human middle ear model, inspired by O'Connor and Puria [7], is integrated with the nonlinear MEA model. The passive BM displacement predicted by the model is compared with the in vivo physiological measurements for the BM displacement in human cadavers [8], and the predicted active BM tuning curves are compared with human psychophysical tuning curves (PTC) from Moore (1978) [9]. The distortion product otoacoustic emissions (DPOAE), predicted by the model at equal stimulus levels and at different stimulus frequencies, have a maximum at around f2/f1=1.2, agreeing with measured data in the literature [10,11].