Evaluation of feedback reduction techniques in hearing aids based on physical performance measures

Area of Science:

• Audiology
• Acoustics
• Signal Processing

Background:

• Hearing aid feedback cancellation is crucial for audiological device performance.
• Existing techniques vary in effectiveness and stability.
• New algorithms require rigorous physical evaluation.

Purpose of the Study:

• To physically evaluate commercial hearing aid feedback cancellation techniques.
• To assess a novel feedback cancellation algorithm's performance.
• To quantify performance metrics including added stable gain, distortion, and tracking.

Main Methods:

• Physical evaluation using measures for instability, oscillations, and distortion.
• Testing with white noise and spectrally colored signals in static acoustic conditions.
• Assessing performance across different input signals (speech, tonal opera).

Main Results:

• Added stable gains ranged from 3 dB to 26 dB.
• Five of six techniques performed worse with tonal signals than speech.
• Preventing drift improved tonal signal performance but worsened tracking and adaptability.

Conclusions:

• Hearing aid feedback cancellation effectiveness is signal- and condition-dependent.
• Algorithm design choices impact trade-offs between stability and adaptability.
• Reproducibility of test setups is critical for reliable evaluation.