Affiliations 

  • 1 Universiti Kebangsaan Malaysia, Faculty of Health Sciences, Centre for Rehabilitation and Special Needs, Audiology Program, Kuala Lumpur, Malaysia. foongyen.chong@ukm.edu.my
  • 2 The University of British Columbia, Faculty of Medicine, School of Audiology & Speech Sciences, Vancouver, Canada
Med J Malaysia, 2018 12;73(6):365-370.
PMID: 30647205

Abstract

INTRODUCTION: Modulation-based noise reduction (MBNR) is one of the common noise reduction methods used in hearing aids. Gain reduction in high frequency bands may occur for some implementations of MBNR and fricatives might be susceptible to alteration, given the high frequency components in fricative noise. The main objective of this study is to quantify the acoustic effect of MBNR on /s, z/.

METHODS: Speech-and-noise signals were presented to, and recorded from, six hearing aids mounted on a head and torso simulator. Test stimuli were nonsense words mixed with pink, cafeteria, or speech-modulated noise at 0 dB SNR. Fricatives /s, z/ were extracted from the recordings for analysis.

RESULTS: Analysis of the noise confirmed that MBNR in all hearing aids was activated for the recordings. More than 1.0 dB of acoustic change occurred to /s, z/ when MBNR was turned on in four out of the six hearing aids in the pink and cafeteria noise conditions. The acoustics of /s, z/ by female talkers were affected more than male talkers. Significant relationships between amount of noise reduction and acoustic change of /s, z/ were found. Amount of noise reduction accounts for 42.8% and 16.8% of the variability in acoustic change for /s/ and /z/ respectively.

CONCLUSION: Some clinically-available implementations of MBNR have measurable effects on the acoustics of fricatives. Possible implications for speech perception are discussed.

* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.