Rodent Auditory Model Bridges Human Hearing Loss Research

A team of researchers led by Morgan Thienpont has developed computational auditory periphery models adapted from humans to mice and gerbils, allowing direct cross-species comparisons for sensorineural hearing loss (SNHL) research. Published as arXiv:2605.19070, the work modifies a 1-D nonlinear cochlear transmission-line model originally designed for humans. By adjusting species-specific parameters—including basilar membrane length and width, stapes area, middle-ear transfer functions, and frequency range—the model captures each species' unique auditory periphery and hearing range. Additional cochlear parameters were calibrated to reproduce realistic tuning and compression. The adapted models for mouse and gerbil were validated against experimental measurements, including basilar membrane velocity level-growth, auditory-nerve tuning curves, and distortion product otoacoustic emissions (DPOAEs). Simulated auditory-nerve outputs matched empirical data for thresholds and frequency selectivity, though discrepancies increased near the cochlear base and apex.

The models successfully reproduced cochlear synaptopathy effects, showing differences in auditory brainstem responses and envelope following responses between mice and gerbils with SNHL. However, individualizing the mouse model's outer hair cell (OHC) damage based on DPOAEs failed to faithfully replicate individual measurements—only intergroup differences in OHC damage were captured. This suggests limitations in using DPOAEs for personalized modeling. Despite this, the framework demonstrates that biophysically grounded auditory models can be translated across species while preserving realistic sound-coding and pathophysiological alterations. The work provides a unified computational tool for bridging animal experiments and non-invasive human diagnostics, potentially accelerating hearing loss research and treatment development.