Computational modeling of the auditory brainstem response to continuous speech.

Area of Science:

• Auditory Neuroscience
• Computational Auditory Neuroscience
• Speech Processing

Background:

• The auditory brainstem response (ABR) is a non-invasive clinical tool for assessing hearing function.
• Previous research demonstrated ABR measurement to continuous speech and its modulation by attention.
• The precise neural origins and signal processing contributing to speech-evoked ABR remain unclear.

Purpose of the Study:

• To elucidate the influence of different brainstem components and speech signal factors on the scalp-recorded ABR.
• To utilize a computational model to simulate and understand the neural processing of speech in the auditory brainstem.

Main Methods:

• Developed a computational model of the auditory brainstem, integrating middle/inner ear, cochlear nuclei, and inferior colliculus models.
• Simulated the neural response to continuous speech at various brainstem stages.
• Validated model simulations against scalp-recorded ABR data from human volunteers.

Main Results:

• Auditory nerve fibers, cochlear nuclei, and inferior colliculus all contribute to speech-evoked ABR.
• The inferior colliculus provides the dominant contribution to the response.
• A wide range of speech harmonics (up to 8 kHz) influence the ABR, with response declining at higher fundamental frequencies.

Conclusions:

• Scalp-recorded speech-evoked ABR predominantly originates in the inferior colliculus.
• The ABR is significantly shaped by numerous speech harmonics, indicating complex, nonlinear auditory processing.
• The model accurately replicates experimental findings, providing insights into auditory brainstem function.