The origin of the distortion product otoacoustic emission fine structure

Distortion-product otoacoustic emissions (DPOAEs) are sounds detected in the ear canal which are generated by the nonlinear processes in the inner ear (cochlea) in response to the external stimulation of two or more tones (primaries). Their generation region in the cochlea can be systematically changed by varying the primary frequencies, and they are currently being evaluated for possible clinical use in screening for hearing defects. The phase and amplitude of various orders of DPOAEs of frequencies, $f\sb {dp}=f1-n(f\sb2-f\sb1),\ (n=1,2,\...),$ were measured in human subjects for two-tone stimuli of frequencies $f\sb1$ and $f\sb2$ $(\u3ef\sb1).$ A number of experimental paradigms (fixed primary ratio $f\sb2/f\sb1,$ fixed $f\sb1,$ fixed $f\sb2,$ and fixed $f\sb {dp})$ were used to investigate the nature of peaks and valleys (fine structure) of DPOAEs in their phase and amplitude dependence on the primary frequencies. This fine structure must be taken into account in any potential clinical applications of DPOAEs. The experimental results largely support a model in which the fine structure stems from interference at the base of the cochlea between distortion product (DP) components coming from the primary DPOAE source region (around the $f\sb2$ tonotopic place) and components coming from the DP tonotopic place (via reflection of an apically moving DP wave). The spectral periodicity of the fine structures for several orders of apical DPOAEs corresponds to a tonotopic displacement of about 0.4 mm along the basilar membrane (BM) (0.4 bark). In agreement with the reaction model, this spectral spacing is also characteristic of synchronous evoked and spontaneous otoacoustic emission spectra as well as the microstructure of the hearing threshold. Approximate analytic expressions for the mechanisms which are responsible for the fine structure are used to interpret the data