Related Experiment Video
Updated: Sep 11, 2025
06:04
Systematic Hearing Performance Evaluation Process for Adolescents with Cochlear Implantation at Early Ages
Published on: March 24, 2023
471
DNN-Based Noise Reduction Significantly Improves Bimodal Benefit in Background Noise for Cochlear Implant Users
Courtney Kolberg1, Sarah O Holbert1, Jamie M Bogle1
1Department of Otolaryngology-Head and Neck Surgery, Division of Audiology, Mayo Clinic, Scottsdale, AZ 85259, USA.
Journal of Clinical Medicine
|August 14, 2025
Summary
Deep neural network (DNN) noise reduction in hearing aids significantly improves speech understanding for bimodal cochlear implant (CI) users. This technology offers a promising solution for CI users struggling with hearing in noisy environments.
Area of Science:
- Audiology
- Neuroscience
- Biomedical Engineering
Background:
- Traditional hearing aid noise reduction offers limited benefit for bimodal cochlear implant (CI) users.
- Deep neural networks (DNNs) have shown promise in improving hearing aid performance in noise.
- This study explores the application of DNNs in hearing aids for bimodal CI users.
Purpose of the Study:
- To investigate the effectiveness of DNN-based noise reduction in hearing aids for bimodal CI users.
- To compare DNN noise reduction performance against traditional noise reduction algorithms in bimodal CI users.
- To assess the impact of DNN technology on speech perception in noise for this population.
Main Methods:
- Eleven bimodal CI users (aged 71-89) were fitted with a hearing aid (HA) featuring DNN-based noise reduction.
- Speech recognition was measured using AzBio sentences in quiet and noise conditions.
- Performance was evaluated with CI alone, HA alone (Calm Situation vs. DNN program), and bimodally.
Main Results:
- The DNN hearing aid program significantly improved bimodal speech recognition in noise, achieving 79% compared to 60% with the Calm Situation program (19% average benefit, p < 0.001).
- In multi-talker babble, the DNN HA program provided a 40% bimodal benefit over CI-alone, significantly outperforming the 21% benefit from the Calm Situation program.
- DNN-based noise reduction demonstrated substantial improvements in speech understanding for bimodal CI users.
Conclusions:
- DNN-based noise reduction in hearing aids significantly enhances speech understanding in noisy environments for bimodal CI users.
- This technology represents a promising advancement for addressing the common challenge of speech perception difficulties in noise faced by bimodal CI users.
- The findings support the clinical utility of DNN noise reduction as a valuable tool for improving the quality of life for bimodal CI recipients.
