Elsevier

Clinical Neurophysiology

Volume 115, Issue 9, September 2004, Pages 2021-2030
Clinical Neurophysiology

Brainstem responses to speech syllables

https://doi.org/10.1016/j.clinph.2004.04.003Get rights and content

Abstract

Objective: http://www.communication.northwestern.edu/csd/research/brainvoltsTo establish reliable procedures and normative values to quantify brainstem encoding of speech sounds.

Methods: Auditory brainstem responses to speech syllables presented in quiet and in background noise were obtained from 38 normal children. Brainstem responses consist of transient and sustained, periodic components—much like the speech signal itself. Transient peak responses were analyzed with measures of latency, amplitude, area, and slope. Magnitude of sustained, periodic frequency-following responses was assessed with root mean square, fundamental frequency, and first formant amplitudes; timing was assessed by stimulus-to-response and quiet-to-noise inter-response correlations.

Results: Measures of transient and sustained components of the brainstem response to speech syllables were reliably obtained with high test–retest stability and low variability across subjects. All components of the brainstem response were robust in quiet. Background noise disrupted the transient responses whereas the sustained response was more resistant to the deleterious effects of noise.

Conclusions: The speech-evoked brainstem response faithfully reflects many acoustic properties of the speech signal. Procedures to quantitatively describe it have been developed.

Significance: Accurate and precise manifestation of stimulus timing at the auditory brainstem is a hallmark of the normal perceptual system. The brainstem response to speech sounds provides a mechanism for understanding the neural bases of normal and deficient attention-independent auditory function.

Introduction

The neural encoding of sound begins in the auditory nerve and travels to the auditory brainstem. Brainstem responses to simple stimuli (e.g., clicks, tones) are widely used in clinical practice in the evaluation of auditory pathway integrity (Møller, 1999, Starr and Don, 1988). Less well-defined is how the brainstem responds to complex stimuli. Describing auditory encoding of speech sounds provides insight into some of the central auditory processes involved in normal communication. Furthermore, this knowledge may be applied to understanding effects of the aging process on hearing, as well as to a broad range of other circumstances, including hearing and communication in individuals with learning problems, peripheral hearing impairments, cochlear implants, or auditory neuropathies.

Some people have normal peripheral hearing, but still cannot perceive speech well. Previous studies have shown that the disruption of neural timing at the cortex is linked to auditory perceptual deficits (Kraus et al., 1996, Tonnquist-Uhlen, 1996, Najaran et al., 1999, Wible et al., 2002). In addition, abnormal electrophysiological responses to speech syllables at the brainstem level have been associated with a wide spectrum of diagnosed learning problems (King et al., 2002; Wible et al., in press). These abnormalities include a temporally delayed response to the onset of a consonant and deficient spectral representation of harmonic aspects of the speech signal. Disruptions of neural encoding in both the brainstem and cortex were exacerbated when speech was presented in background noise (Cunningham et al., 2001).

Part of the difficulty in perceiving consonants in noisy situations is that they are rapid, relatively low-amplitude transient features of speech. Stop consonants, such as /d/, are known to be particularly vulnerable to disruption by background noise in normal and clinical populations (Brandt and Rosen, 1980). The perception of vowels, however, is more resistant to the effects of noise because they are periodic, sustained signals, and generally louder than consonants.

Brainstem responses provide direct information about how the sound structure of a speech syllable is encoded by the auditory system. It is particularly compelling to consider that specific aspects of the sound structure of the acoustic signal are maintained and reflected in the neural code. Similar to the speech syllable itself, the brainstem response to a speech syllable can be divided into transient and sustained portions, namely the onset response and the frequency-following response (FFR) (Boston and Møller, 1985). Onset responses are transient, with peak durations lasting tenths of milliseconds, thus we will refer to these rapid deflections as transient responses. Within the FFR are discrete peaks corresponding to the periodic peaks in the stimulus waveform. However, this region can be considered as a whole, as it contains a periodic signal sustained for tens or hundreds of milliseconds. Although peaks within the FFR may be thought of as successive onsets, for descriptive purposes, we will use the term FFR to refer to the later portion of the response evoked by the harmonic vowel structure of the stimulus. There is a parallel effect of noise on the brainstem response, similar to the disruption of speech perception, in that transient onsets were more affected by the noise, sometimes even eliminated, while the sustained portion remained intact (Cunningham et al., 2001).

The specific aims of this study were: (1) to delineate measures of the timing and magnitude of the brainstem response to the speech syllable /da/ in quiet and background noise; (2) to establish normative values for these features; and (3) to determine the test–retest reliability of these measures.

Section snippets

Subjects

Thirty-eight children, ages 8–12 years (21 male, 17 female) participated in the primary focus of this study, which established normative values for the brainstem response to speech syllables. Eight children (four male, four female) were part of the retest reliability portion of the study. None of the children had a history of medical or learning problems and all performed within normal limits on laboratory-internal standardized measures of learning and academic achievement. These measures

Results

Based on our evaluation of 38 subjects' responses recorded in quiet and 36 subjects' responses recorded in background noise, normative values for the aforementioned brainstem measures were established. Table 2 shows means and SDs for discrete peak measures obtained in quiet and background noise. Table 3, Table 4 provide timing and magnitude values, respectively, for the FFR.

Discussion

The ability to quantify a brainstem response elicited by speech sounds provides a powerful tool for research and clinical use. The speech-evoked brainstem response faithfully reflects many acoustic properties of the speech signal. In the normally perceiving auditory system, stimulus timing, on the order of fractions of milliseconds, is accurately and precisely represented at the level of the brainstem. Overall, the brainstem response provides a mechanism for understanding the neural bases of

Conclusions

Brainstem response timing and magnitude measures provide reliable information about the neural encoding of speech sounds. This study outlined specific measures of brainstem function that may be used to characterize neural encoding of speech sounds for clinical and research applications. Transient and sustained measures provide information regarding auditory pathway encoding of brief and periodic aspects of the stimulus. Some of the data suggest that transient and sustained responses represent

Acknowledgements

The National Institute of Health NIDCD R01-01510 supported this research. We thank the children and their families who participated in this study. We also thank Steven Zecker, for statistical consultation; Erika Skoe, for the development of programs to analyze the data; and also members of the Kraus laboratory who tested subjects and gave their support throughout the study.

References (19)

There are more references available in the full text version of this article.

Cited by (0)

View full text