Behavioral assessment of CID's benchtop (version 2) digital hearing aid: noise reduction

Area of Science:

• Audiology
• Speech processing
• Hearing aid technology

Background:

• Digital hearing aids utilize noise reduction algorithms to improve speech intelligibility in noisy environments.
• Adaptive and nonadaptive algorithms aim to reduce background noise, but their efficacy can vary.
• Understanding how hearing-impaired individuals perceive and select these algorithms is crucial for device optimization.

Purpose of the Study:

• To evaluate the effectiveness of various digital hearing aid noise reduction algorithm parameters.
• To determine if perceptual ratings align with optimized word intelligibility.
• To investigate the influence of speech input level and noise type on algorithm performance.

Main Methods:

• Employed perceptual rating and word identification tests.
• Recruited 15 adults with hearing impairments.
• Tested speech intelligibility at soft, moderate, and loud levels in quiet, fan noise, and multitalker babble.

Main Results:

• Hearing-impaired listeners' clarity judgments successfully identified algorithms that enhanced word intelligibility at soft and moderate speech levels.
• The spectral adaptation requirements of the algorithms were not consistent across different signal levels.
• Performance varied depending on the type of noise present.

Conclusions:

• Perceptual feedback from hearing-impaired users is valuable for selecting effective noise reduction strategies.
• Current adaptive algorithms may require adjustments based on speech loudness for optimal performance.
• Further research is needed to refine spectral adaptation for varying signal intensities in hearing aids.