Vowel Acoustics as Predictors of Speech Intelligibility in Dysarthria

Abstract

Purpose: To examine the predictive value of a selection of acoustic vowel measures for predicting intelligibility (i.e., measured using both orthographic transcriptions [OT] and visual analog scale [VAS] ratings) in speakers with dysarthria. The following questions were posed: (1) How well do trajectory-based and token-based vowel space measures predict intelligibility? And (2) does the relationship between vowel measures and intelligibility differ based on the type of intelligibility measurement (i.e., OT vs. VAS ratings)?Method: The Grandfather Passage was read aloud by forty speakers with dysarthria of varying etiologies, including Parkinson's disease (n = 10), amyotrophic lateral sclerosis (n = 10), Huntington's disease (n = 10), and cerebellar ataxia (n = 10). Token-based (i.e., acoustic vowel space area [VSA], corner dispersion) and trajectory-based (i.e., VSA hull area, and vowel space density [VSD]) acoustic vowel measures were calculated. Naïve listeners (N = 140) were recruited via crowdsourcing to provide OT and VAS intelligibility ratings. Hierarchical linear regression models were created to model OT and VAS ratings of intelligibility using the acoustic vowel measures as predictors.Results: Traditional VSA was the sole significant predictor of speech intelligibility for both the OT and VAS models. In contrast, the trajectory-based measures were not significant predictors of intelligibility. Additionally, the OT and VAS intelligibility ratings conveyed similar information.Conclusions: The findings suggest that traditional token-based vowel measures better predict intelligibility than trajectory-based measures. Additionally, the findings suggest that VAS methods are comparable to OT methods for estimating speech intelligibility for research purposes.