Speech Intelligibility and Spatial Release From Masking Improvements Using Spatial Noise Reduction Algorithms in...

Area of Science:

• Audiology
• Speech and Hearing Sciences
• Biomedical Engineering

Background:

• Bimodal cochlear implant (CI) users, those with a CI in one ear and a hearing aid (HA) in the other, often struggle with speech understanding in noisy environments.
• Spatial noise reduction algorithms aim to enhance speech intelligibility by separating speech from background noise.
• The effectiveness of different beamformer algorithms in bimodal CI users requires further investigation.

Purpose of the Study:

• To evaluate the speech intelligibility benefit of two spatial noise reduction algorithms, adaptive differential microphones (ADM) and minimum variance distortionless response (MVDR), in bimodal CI users.
• To compare the performance of monaural ADM and binaural MVDR algorithms in simulated real-world listening conditions.

Main Methods:

• Ten bimodal CI users with moderate to severe contralateral hearing loss participated.
• A master hearing aid platform (MHA) was used to apply hearing loss compensation and noise reduction algorithms.
• Speech reception thresholds were measured in various noise conditions, including stationary noise and a 20-talker babble, using simulated cafeteria scenarios with controlled head movements.

Main Results:

• Both ADM and MVDR algorithms provided significant improvements in speech reception thresholds for frontal speech in the presence of spatially separated stationary noise.
• MVDR demonstrated greater benefits than ADM, particularly in a challenging 20-talker babble noise scenario (7.5 dB vs. 3.5 dB improvement).
• No benefit was observed when speech and noise were co-located.

Conclusions:

• Beamformer algorithms like ADM and MVDR offer substantial speech intelligibility benefits for bimodal CI users in everyday noisy situations.
• The binaural MVDR algorithm shows superior performance compared to bilaterally applied monaural ADM.
• These findings support the integration of advanced noise reduction strategies into devices for bimodal CI users to improve their listening experience.