Hannah Rowe(Massachusetts General Hospital Institute of Health Professions (MGH IHP)), Sarah Gutz(Harvard University), Marc Maffei(Massachusetts General Hospital Institute of Health Professions (MGH IHP)) and Jordan Green(MGH IHP)
The purpose of this study was to determine the articulatory phenotypes of amyotrophic lateral sclerosis (ALS) and Parkinson's disease (PD) using a novel acoustic-based framework that assesses five key components of motor performance: Coordination, Consistency, Speed, Precision, and Rate. The use of interpretable, hypothesis-driven features has the potential to inform impairment-based automatic speech recognition (ASR) models and improve classification algorithms for disorders with divergent articulatory profiles. Acoustic features were extracted from audio recordings of 18 healthy controls, 18 participants with ALS, and 18 participants with PD producing syllable sequences. Results revealed significantly different articulatory phenotypes for each disorder group. Upon stratification into Early Stage and Late Stage in disease progression, results from individual receiver operating characteristic (ROC) curves and decision tree analyses showed high diagnostic accuracy for impaired Coordination in the Early Stage and impaired Rate in the Late Stage. With additional research, articulatory phenotypes characterized using this framework may lead to advancements in ASR for dysarthric speech and diagnostic accuracy at different disease stages for individuals with distinct articulatory deficits.