Effects of Transient Noise Reduction Algorithms on Speech Intelligibility and Ratings of Hearing Aid Users

Abstract

The goal of this study was to assess the functional utility of transient noise reduction (TNR) algorithms available in hearing aids via speech intelligibility and user preferences. Two pairs of hearing aids, 1 pair each from Siemens and Unitron, were programmed for 17 hearing impaired individuals after a hearing evaluation. Intelligibility was measured for each participant for sentences presented in quiet, with 2 types of transient noise, multitalker babble, and in a combination of each type of transient noise and multitalker babble. Each condition was tested with TNR activated and TNR deactivated in a counterbalanced, single-blinded format. Subjective ratings of overall speech understanding, comfort, and sound quality were obtained for each condition. A significant improvement in speech intelligibility was measured with the TNR activated when speech was presented in multitalker babble, in the presence of chair clang transient noises, and when combining these noises. Activation of the TNR algorithm did not result in significant improvements for any of the subjective ratings. While improvements were limited to certain conditions, specifically those with the chair clang transient and/or multitalker babble, TNR appears to offer an incremental step in improving the listening experience for hearing aid users.