Spectrotemporal modulation sensitivity as a predictor of speech intelligibility in noise with hearing aids

• The audiogram predicts less than a third of the variance in speech reception thresholds (SRTs) for hearing-impaired (HI) listeners properly fit with individualized frequency-dependent gain. The remaining variance is often attributed to a combination of su-prathreshold distortion in the auditory pathway and non-auditory factors such as cogni-tive processing. Distinguishing between these factors requires a measure of suprathresh-old auditory processing to account for the non-cognitive contributions. Preliminary re-sults in 12 HI listeners identified a correlation between spectrotemporal modulation (STM) sensitivity and speech intelligibility in noise presented over headphones. The cur-IHCON 2014 27 August 13-17, 2014rent study assessed the effectiveness of STM sensitivity as a measure of suprathreshold auditory function to predict free-field SRTs in noise for a larger group of 47 HI listeners with hearing aids.SRTs were measured for Hagerman sentences presented at 65 dB SPL in stationary speech-weighted noise or four-talker babble. Pre-recorded speech and masker stimuli were played through a small anechoic chamber equipped with a master hearing aid pro-grammed with individualized gain. The output from an IEC711 Ear Simulator was played binaurally through insert earphones. Three processing algorithms were examined: linear gain, linear gain plus noise reduction, or fast-acting compressive gain.STM stimuli consist of spectrally-rippled noise with spectral-peak frequencies that shift over time. STM with a 2-cycle/octave spectral-ripple density and a 4-Hz modulation rate was applied to a 2-kHz lowpass-filtered pink-noise carrier. Stimuli were presented over headphones at 80 dB SPL (±5-dB roving). The threshold modulation depth was estimated adaptively in a two-alternative forced-choice task.STM sensitivity was strongly correlated (R2=0.48) with the global SRT (i.e., the SRTs averaged across masker and processing conditions). The high-frequency pure-tone aver-age (3-8 kHz) and age together accounted for 23% of the variance in global SRT. STM sensitivity accounted for an additional 28% of the variance in global SRT (total R2=0.51) when combined with these two other metrics in a multiple-regression analysis. Correla-tions between STM sensitivity and SRTs for individual conditions were weaker for noise reduction than for the other algorithms, and marginally stronger for babble than for sta-tionary noise.The results are discussed in the context of previous work suggesting that STM sensitivity for low rates and low carrier frequencies is impaired by a reduced ability to use temporal fine-structure information to detect slowly shifting spectral peaks. STM detection is a fast, simple test of suprathreshold auditory function that accounts for a substantial pro-portion of variability in hearing-aid outcomes for speech perception in noise.

Ämnesord

SAMHÄLLSVETENSKAP  -- Annan samhällsvetenskap (hsv//swe)

SOCIAL SCIENCES  -- Other Social Sciences (hsv//eng)

Publikations- och innehållstyp

ref (ämneskategori)

kon (ämneskategori)