A Retrospective Study of Acoustic Measures of Glottal Stop Production to Assess Vocal Function in Unilateral Vocal Fold Paresis/Paralysis Patients

Abstract

Purpose: The aim of this study was to determine (a) diagnostic accuracy of acoustic measures of glottal stop production (GSP; intensity differences, slopes, complete voicing cessation) to distinguish between unilateral vocal fold paresis/paralysis (UVFP) patients and controls; (b) if acoustic measures of GSP significantly correlated with an acoustic measure of voice disorder severity, acoustic voice quality index (AVQI); and (c) if acoustic measures from another type of voicing cessation, voiceless consonant production, also significantly differed between groups. Method: Ninety-seven patients with unilateral paresis/paralysis and 35 controls with normal laryngostroboscopic signs produced two sets of five repeated [i] and four repeated [isi]. Tokens were randomized by type between groups and analyzed blinded using a customized Praat program that computed intensity differences and slopes between vowel maxima and glottal stop minima for inter-[i] tokens and vowel maxima and voiceless consonant minima for intra-[isi] tokens. The number of voicing cessations for inter-[i] tokens was obtained. Results: Onset and offset intensity differences and number of voicing cessations from inter-[i] tokens had the greatest areas under the curve (.854, .856, and .835, respectively). Correlation coefficients were significant ( p < .01) between AVQI and all GSP acoustic measures with weak/medium effect sizes. No significant differences were found between controls and participants with UVFP for acoustic measures from intra-[isi]. Conclusions: Acoustic GSP measures demonstrated good diagnostic accuracy and some relationship to severity of voice disorder. No significant differences in acoustic measures for medial voiceless fricative consonants between controls and participants with UVFP suggested that voicing cessation for voiceless fricatives differs from voicing cessation for GSP.