Predicting effects of hearing-instrument signal processing on consonant perception

Area of Science:

• Auditory Neuroscience
• Speech Processing
• Acoustics

Background:

• Hearing aids (HA) and cochlear implants (CI) employ signal processing that can alter speech perception.
• Understanding these alterations is crucial for optimizing hearing instrument performance.

Purpose of the Study:

• To investigate the impact of HA and CI processing on consonant perception in normal-hearing listeners.
• To validate a speech perception model's ability to predict these effects.

Main Methods:

• Measured consonant perception in normal-hearing listeners with simulated HA processing (frequency compression, noise suppression).
• Utilized existing consonant perception data from normal-hearing listeners with CI processing (noise-vocoded stimuli).
• Compared experimental data with predictions from a speech perception model combining auditory processing and template matching.

Main Results:

• The speech perception model showed high agreement with perceptual data for both HA and CI simulations.
• The model accurately predicted consonant recognition and confusion patterns.
• This demonstrates the model's sensitivity to supra-threshold processing effects.

Conclusions:

• The validated speech perception model can effectively simulate the effects of HA and CI processing on consonant perception.
• This model is a valuable tool for evaluating and developing hearing instrument signal processing strategies.
• Findings can inform personalized hearing loss simulations and aid in audiological rehabilitation.