A Signal-to-Noise Ratio Model for the Speech-Reception Threshold of the Hearing Impaired

TLDR

Speech‑reception threshold (SRT) modeling is essential for understanding how hearing‑impaired listeners perceive speech in noise. This paper reviews investigations that apply a quantitative SRT model to characterize hearing‑impaired performance. The model incorporates six parameters—two for normal‑hearing SRT, two for hearing‑loss attenuation and threshold elevation, and three for hearing‑aid acoustic gain, threshold elevation, and internal noise—and is validated with data from presbycusis, pathological, and noise‑induced losses. The model accurately fits the data and shows that many hearing‑impaired individuals struggle with speech intelligibility at noise levels above 50–60 dB(A), while hearing aids are most effective below that level.

Abstract

This paper reviews the results of a series of investigations inspired by a model of the speech-reception threshold (SRT) of hearing-impaired listeners. The model contains two parameters accounting for the SRT of normal-hearing listeners (SRT in quiet and signal-to-noise ratio corresponding to the threshold at high noise levels), two parameters describing the hearing loss (attenuation and threshold elevation in terms of signal-to-noise ratio), and three parameters describing the hearing aid (acoustic gain, threshold elevation expressed in signal-to-noise ratio, and equivalent internal noise level). Experimental data are reported for three different types of hearing impairment: presbycusis, hearing losses with a pathological origin, and noise-induced losses. The model gives an excellent description of the data. It demonstrates that for many hearing-impaired persons speech intelligibility at noise levels beyond 50 to 60 dB(A) is their main problem, whereas hearing aids are most effective below that noise level.