The effect of a linked bilateral noise reduction processing on speech in noise performance

Directional processing already provides tangible noise reduction benefits in hearing aids but further improvement is needed for hearing-impaired listeners to communicate as effectively as normal-hearing listeners in noisy environments. The objective of this study was to investigate if a binaurally linked beamformer could further improve the signal-to-noise ratio (SNR). Speech reception thresholds (SRT) and spatial perception were compared for bilaterally fitted cardioid microphones and two binaurally linked beamformer processing conditions; 1) a single audio stream output to the two ears, and 2) two audio stream outputs which preserved spatial cues. 10 normal-hearing and 22 hearing-impaired listeners were recruited for this study. The strategies were implemented on a real-time PC processing platform, wired to a pair of behind-the-ear devices via a sound interface. A speech-in-noise test was administered using the Bamford-Kowal-Bench (BKB) sentences targeting the SNR for which 75% correct keywords were identified in spatially separated multi-talker babble noise and room reverberation. The SNR level at which the listeners acquired 95% intelligibility from continues speech discourse material, using a male and a female talker, was also obtained. Sound amplification was provided according to NAL-NL2. Both beamformer conditions improved the SRTs relative to the conventional cardioids, but by a greater degree for the hearing-impaired listeners, and more convincingly at the higher SRTs.