Impact of Reduced Spectral Resolution on Temporal-Coherence-Based Source Segregation

Area of Science:

• Auditory Neuroscience
• Speech Perception
• Hearing Impairment

Background:

• Hearing-impaired listeners, including those with cochlear implants (CIs), face challenges understanding speech in noisy environments.
• Effective listening in noise relies on the brain's auditory source segregation, grouping sounds into distinct perceptual streams.
• Temporal coherence across frequency channels aids normal-hearing listeners in source segregation.

Approach:

• Investigated the hypothesis that reduced spectral resolution impairs temporal-coherence-based auditory segregation.
• Utilized a physiologically plausible computational model of temporal-coherence processing.
• Simulated CI listening conditions and varying degrees of sensorineural hearing loss.

Key Points:

• Reduced spectral resolution, common in hearing loss and CI use, hinders the brain's ability to segregate sound sources based on temporal coherence.
• Computational model predictions align with behavioral data, showing CI current spread reduces comodulation masking release (CMR) and speech intelligibility.
• Model predicts diminished CMR with increased outer-hair-cell damage, indicating a broader impact of spectral degradation.

Conclusions:

• Reduced spectral resolution is a key factor impairing auditory segregation and speech-in-noise understanding for hearing-impaired listeners.
• Findings highlight the detrimental effect of spectral smearing on temporal-coherence processing, crucial for complex auditory scenes.
• The study provides insights into the mechanisms underlying listening difficulties in sensorineural hearing loss and CI users.