Initial development of a temporal-envelope-preserving nonlinear hearing aid prescription using a genetic algorithm.

Area of Science:

• Audiology
• Signal Processing
• Biomedical Engineering

Background:

• Current hearing aid prescriptions often neglect the impact of temporal envelope distortions on speech perception.
• Evidence suggests that preserving the temporal envelope shape can improve speech understanding for hearing aid users.

Purpose of the Study:

• To develop and evaluate a novel hearing aid prescription method using a genetic algorithm (GA).
• To assess if the GA-based prescription can preserve the temporal envelope shape of speech.
• To compare the GA prescription's performance against a standard prescription (NAL-NL2) in individuals with hearing impairment.

Main Methods:

• A genetic algorithm (GA) was designed to optimize multiband compression settings.
• The GA aimed to preserve the temporal envelope shape and fit it within the listener's dynamic range.
• Hearing aid prescriptions generated by the GA were compared to the NAL-NL2 prescription using behavioral tests.

Main Results:

• The GA-generated prescription featured a low compression threshold, short attack/release times, and optimized gain/compression ratio.
• No significant difference in speech-in-noise perception was found between the GA and NAL-NL2 prescriptions.
• Gap detection performance was significantly better with the GA prescription compared to NAL-NL2.

Conclusions:

• This study presents a proof of concept for incorporating temporal envelope preservation into hearing aid prescriptions.
• The GA-based approach offers a potential method to reduce speech perception errors caused by temporal distortions.
• Further research is warranted to explore the clinical benefits of this novel prescription strategy.