Allyn E. Hubbard
Dept. of Biomed. Eng.; Otolaryngol. Elect., Comput. and Systems, Boston Univ., 44 Cummington St., Boston, MA 02215
Denise A. Gonzales
David C. Mountain
Boston Univ., Boston, MA 02215
A traveling-wave model of the cochlea [A. E. Hubbard, Science 259, 68--71 (1993)] is comprised of two transmission lines coupled tightly by feedback. The model works because the variable speed of the traveling wave on one transmission line, a classical long-wave model, slows over a region to match approximately the natural propagation velocity of the second line. When this happens, amplification occurs. A fundamental difference between this model and many other active models is that power is generated principally in the tip region of the response, not in the region basal to the tip. The traveling-wave amplifier model replicates several experimental results: stapes/basilar-membrane-velocity ratios versus frequency and the effects of cochlear damage. The original model has been changed so that inter-line coupling function is nonlinear. Basilar-membrane velocity responses having good correspondence with experimental data have been obtained using click pressure stimuli at varying levels. The distortion product emission (DPOE) 2f[sub 1]-f[sub 2] produced by the model shows qualitative correspondence with experimental data. [Work supported by ONR and NIH.]