Objective measures of frequency selectivity.

Abstract

A common difficulty often reported by hearing-impaired listeners is an inability to understand speech in a background of noise. A number of investigations have suggested that a possible explanation for speech understanding difficulties, especially in noise, is impaired frequency selectivity. Unfortunately, the length and complexity of current psychophysical methodologies preclude their widespread utilization in clinical environments. Investigations that show a reliable correlation between electrophysiologic measures and critical bandwidth [S. Zerlin, J. Acoust. Soc. Am. 79, 1612–1616 (1986)] may overcome these limitations. However, additional researchers have not confirmed these relationships [D. L. Burrows and S. J. Barry, J. Acoust. Soc. Am. 88, 180–184 (1990)]. Because the concept of reliable electrophysiological measures of frequency selectivity remains uncertain, the present investigation examined electrophysiologic measures of frequency selectivity by recording the amplitude of wave V of the auditory brainstem response as a function of the frequency separation of two components of a tonal complex. Seven young adult listeners with normal hearing served as subjects. The frequency separation of the two component tones in the complex was varied as a function of center frequency from 100–500 Hz in steps of 100 Hz around a center frequency of 2 kHz and from 400–1200 Hz in steps of 200 Hz around a center frequency of 4 kHz. The width of the critical band was defined as the smallest amount of separation of the two-tone complex that exhibited a reliable increase in the amplitude of wave V. Results indicate significant enhancements in wave V as a function of the frequency separation of the tonal complex. The critical bandwidths measured, however, were slightly larger than those reported in psychophysical studies. The implications for clinical measures of frequency selectivity in normal and hearing impaired listeners will be discussed.