RSS-Feed abonnieren
DOI: 10.3766/jaaa.25.7.5
Directional Benefit Is Present with Audiovisual Stimuli: Limiting Ceiling Effects
Publikationsverlauf
Publikationsdatum:
06. August 2020 (online)
Background: Past research demonstrates that as the speech recognition performance of listeners fitted with omnidirectional hearing aids approaches maximum (ceiling), the benefit afforded by directional microphones is necessarily lessened. This effect could potentially eliminate the benefit provided by directional microphones in easier listening situations, such as environments with visual cues and favorable signal-to-noise ratios (SNRs).
Purpose: The purpose of this study was to evaluate directional benefit using auditory-visual stimuli in SNRs commonly found in the real world (ranging from +3 to +12 dB). In order to maximize the possibility of directional benefit, a bilateral beamformer was used, which functions as a highly directional microphone.
Research Design: Sentence recognition was evaluated in three SNRs and in two levels of reverberation (low and moderate). For each SNR and reverberation combination, sentence recognition was evaluated using omnidirectional and bilateral beamformer microphone modes.
Study Sample: A total of 15 adults with hearing loss participated. Only listeners who had significant difficulty understanding speech in noise were included.
Data Collection and Analysis: Sentence recognition scores were evaluated using analysis of variance with three within-participant variables (SNR, reverb, microphone mode). Follow-up analyses were conducted using linear contrast while controlling for family-wise error rate.
Results: Results revealed significant bilateral beamformer benefit ranging from approximately 22–30 rationalized arcsine units (˜20–28 percentage points) in both low and moderate reverberation across all tested SNRs (+3 to +12 dB).
Conclusions: These results provide support for the use of bilateral beamformers, even at relatively favorable SNRs and in the presence of visual cues, for listeners who demonstrate poor sentence recognition in noise.