Background: The hearing of tetrapods including humans is enhanced by an active process that amplifies the mechanical inputs associated with sound, sharpens frequency selectivity, and compresses the range of responsiveness. The most striking manifestation of the active process is spontaneous otoacoustic emission, the unprovoked emergence of sound from an ear. Hair cells, the sensory receptors of the inner ear, are known to provide the energy for such emissions; it is unclear, though, how ensembles of such cells collude to power observable emissions.
Methodology and principal findings: We have measured and modeled spontaneous otoacoustic emissions from the ear of the tokay gecko, a convenient experimental subject that produces robust emissions. Using a van der Pol formulation to represent each cluster of hair cells within a tonotopic array, we have examined the factors that influence the cooperative interaction between oscillators.
Conclusions and significance: A model that includes viscous interactions between adjacent hair cells fails to produce emissions similar to those observed experimentally. In contrast, elastic coupling yields realistic results, especially if the oscillators near the ends of the array are weakened so as to minimize boundary effects. Introducing stochastic irregularity in the strength of oscillators stabilizes peaks in the spectrum of modeled emissions, further increasing the similarity to the responses of actual ears. Finally, and again in agreement with experimental findings, the inclusion of a pure-tone external stimulus repels the spectral peaks of spontaneous emissions. Our results suggest that elastic coupling between oscillators of slightly differing strength explains several properties of the spontaneous otoacoustic emissions in the gecko.