Category: Technology
Objective: To evaluate the reliability and validity of automatedly generated vowel space metrics from remotely collected smartphone speech samples as biomarkers of articulatory impairment severity and progression in individuals with early Parkinson’s Disease (PD).
Background: PD hypokinetic dysarthria is typified by a reduced range of articulatory motion, quantifiable with vowel space-based acoustic measures e.g. vowel articulation index (VAI), vowel space area (VSA). Although previous studies demonstrated clinical validity of these measures, they relied on manual estimation from speech samples collected in controlled acoustic environments, and existing automatic methods require validation in early stage PD samples.
Method: 316 early-stage, drug-naive PD participants of the PASADENA Phase II study (NCT03100149) performed reading and spontaneous speech tasks every other day at home with the Roche PD Mobile Application v2. VAI and VSA estimates were automatically computed and aggregated over two-week periods. Intra-class correlations (ICC) quantified test-retest reliabilities. Baseline sex differences were tested with ANCOVAs, clinical validity with Spearman correlations with the bulbar score (sum of MDS-UPDRS items 2.1, 2.2, 2.3, 3.1, 3.2)) and progression sensitivity with linear mixed effect models fitted on feature changes from baseline to week 52. Progression sensitivity was evaluated in placebo patients only (n=94).
Results: VAI and VSA evidenced good test-retest reliability (both ICCs>=.75).
At baseline, females evidenced higher VSA and VAI (p<.001).
Both features correlated with bulbar scores (VSA, M: r=-.25(p<.001), F: r=-.26(p<.01); VAI, M: r=-.14(p<.05), F: r=-.21(p<.05)) at baseline. Both features worsened over 1 year (VSA, p<.05;VAI, p<.01), independent of sex (sex:fortnight, VSA, p=.569;VAI, p=.784).
Conclusion: Automatedly generated vowel space metrics from remotely collected smartphone speech samples may represent reliable and valid biomarkers of dysarthria severity and progression in individuals with early-stage PD.
To cite this abstract in AMA style:
D. Kwaśny, J. Rusz, R. Ullmann, E. Volkova-Volkmar, B. Fehlmann, S. Lambrecht, G. Pagano, T. Kustermann, R. Postuma, M. Lindemann, K. Taylor. Reliability and validity of automatedly generated vowel space metrics as biomarkers of articulatory impairment severity and progression in early Parkinson’s Disease [abstract]. Mov Disord. 2023; 38 (suppl 1). https://www.mdsabstracts.org/abstract/reliability-and-validity-of-automatedly-generated-vowel-space-metrics-as-biomarkers-of-articulatory-impairment-severity-and-progression-in-early-parkinsons-disease/. Accessed June 15, 2025.« Back to 2023 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/reliability-and-validity-of-automatedly-generated-vowel-space-metrics-as-biomarkers-of-articulatory-impairment-severity-and-progression-in-early-parkinsons-disease/