Symmetrical and asymmetrical directional benefits are present for talkers at the front and side

Abstract

Objective

The purpose of the study was to examine the effects of symmetrical and asymmetrical directional microphone settings on speech recognition, localisation and microphone preference in listening conditions with on- and off-axis talkers.

Design

A within-subjects repeated-measure evaluation of three hearing aid microphone settings (bilateral omnidirectional, bilateral directional, asymmetrical directional) was completed in a moderately reverberant laboratory. An exploratory analysis of the potential relationship between microphone preference and unaided measures was also completed.

Study Sample

Twenty adult listeners with mild to moderately severe bilateral hearing loss participated.

Results

The directional and asymmetric microphone settings resulted in equivalent benefits for sentence recognition in noise, word recall, and localisation speed regardless of the speech loudspeaker location (on- or off-axis). However, localisation accuracy was significantly worse with the asymmetric fitting than the directional setting when speech was presented from the rear hemisphere. Listeners who always preferred directional microphones had significantly poorer unaided speech recognition than those who preferred the omnidirectional setting for one or more listening condition.

Conclusions

Benefits from directional and asymmetric processing were small in the current study, but generally similar to each other. Unaided speech recognition in noise performance may have utility as a clinical predictor of preference for directional processing.

Keywords:

• Hearing aids
• directional benefit
• microphones
• localisation
• speech recognition

Acknowledgements

The authors would like to thank Kjersten Branscome, Amy Stahl and Haiping Huang for their assistance with participant recruitment and data collection.

Disclosure statement

The authors declare that there is no conflict of interest.

Additional information

Funding

This research was funded by a grant from GN Resound (Todd Ricketts, PI) and the Maddox Charitable Trust.