Related Experiment Videos
Output signal-to-noise ratio and speech perception in noise: effects of algorithm
Christi W Miller1, Ruth A Bentler2, Yu-Hsiang Wu2
1a Department of Speech and Hearing Sciences , University of Washington , Seattle , WA , USA.
International Journal of Audiology
|March 31, 2017
Summary
Hearing aid (HA) processing changed signal-to-noise ratio (SNR) significantly, but these improvements did not predict better speech perception in noise for individuals with hearing loss.
Area of Science:
- Audiology
- Signal Processing
- Hearing Science
Background:
- Hearing aids (HAs) utilize compression and noise reduction (NR) to improve audibility.
- The impact of these processing strategies on signal-to-noise ratio (SNR) and subsequent speech perception is not fully understood.
Purpose of the Study:
- To quantify SNR changes from compression and NR in HAs from different manufacturers.
- To assess if these SNR changes predict speech perception improvements in individuals with normal and impaired hearing.
Main Methods:
- Quantified output SNR using a phase-inversion technique across three HA manufacturers.
- Employed a linear mixed model to correlate SNR changes with aided speech perception in noise.
- Included participants with normal hearing and mild to moderately severe sensorineural hearing loss.
Main Results:
- HA processing resulted in small but statistically significant SNR changes.
- Significant interactions were found between HA devices and processing types regarding SNR.
- Crucially, the measured changes in SNR did not predict improvements in speech perception.
Conclusions:
- While HA compression and NR algorithms alter SNR, these changes do not directly translate to enhanced speech perception.
- The functional benefit of these algorithms may lie beyond direct SNR improvement for speech understanding.