In response a two-tone stimulus of frequencies f1 and f2, the cochlea emits distortion product otoacoutic emissions (DPOAEs) of frequencies n1 f1 +n2 f2 where n1 and n2 are signed integers. The commonly studied DPOAE is the 2f1-f2 DPOAE – but the 2f2-f1 DPOAE has also been reported in some experiments. However, the 2f1-f2 and 2f2-f1 are commonly hypothesized to be generated by different mechanisms. While the 2f1-f2 is primarily generated due to nonlinear distortion at the locations where the response to the two primary tones interact, the 2f2-f1 is primarily generated due to inhomogeneities on the basilar membrane that reflects traveling waves (Fig. 1).
In this work, we used a 3D physiologically-motivated computational model of the cochlea to simulate both the 2f1-f2 and 2f2-f1 DPOAEs. To model the reflection source, random inhomogeneities were added to the properties of the outer hair cell. The total DPOAE was decomposed into distortion source and reflection source components. We found that, as in typically experiments, the 2f1-f2 DPOAE is dominated by the distortion source, except for low primary frequency ratio and high sound pressure level; the 2f2-f1 DPOAE is dominated by the reflection source for all stimulus parameters.
The use of a computational model made it possible to gain considerable insight into how these two emissions propagate. For example, for the 2f2-f1 DPOAE, both distortion and reflection components are generated inside the cochlea; however, only the reflection component is able to propagate towards the stapes.
More information about this work can be found in our paper:
Wen, H., Bowling, T., Meaud, J., 2018, Investigation of the 2f1-f2 and 2f2-f1 distortion product otoacoustic emissions using a computational model of the gerbil ear, Hearing Research, 365:127-140