On the role of auditory nerve fiber types in loudness encoding and perception in humans

It is unclear how loudness is encoded in the auditory system. In this study, we investigated how place-dependent ratios of high-, medium-, and low-spontaneous rate fibers can be optimized for loudness perception using a computational model. In our simulations, the total spike count of the auditory nerve (AN), in response to 1-s tones, was obtained using the Bruce etal.'s (2018) model. The model was adapted to apply a KEMAR head-related transfer function to stimulus inputs. Equal spike contours (ESC) were compared to traditional equal loudness contours (ELC; ISO 226, 2003). Optimization algorithms were used to tune the local ratio of the fiber types as a function of cochlear place to minimize errors between ELC and ESC. Our preliminary results show that the distribution of fiber types varies along the length of the cochlea, improving the ELC-ESC match compared to a uniform distribution. This distribution may reflect the actual ratio of spontaneous rate fibers present in the human cochlea. Simulating cochlear nerve degeneration in the model elicits changes in ESC, potentially leading to new diagnostic methods for the condition. [Work supported by grant R01-DC07910 from the NIDCD of NIH.]