Speech-based identification of L-DOPA ON/OFF state in Parkinson’s Disease subjects

Abstract

BackgroundParkinsons disease patients (PDP) are evaluated using the unified Parkinsons disease rating scale (UP-DRS) to follow the longitudinal course of the disease. UP-DRS evaluation is performed by a neurologist, and hence its use is limited in the evaluation of short-term (daily) fluctuations. Subjects taking L-DOPA as part of treatment to reduce symptoms exhibit motor fluctuations as a common complication.\n\nObjectivesThe aim of the study is to assess the use of speech analysis as a proxy to continuously monitor PDP medication state.\n\nMethodsWe combine acoustic, prosody, and semantic features to characterize three speech tasks (picture description, reverse counting and diadochokinetic rate) of 25 PDP evaluated under different medication states: \"ON\" and \"OFF\" L-DOPA.\n\nResultsClassification of medication states using features extracted from audio recordings results in cross-validated accuracy rates of 0.88, 0.84 and 0.71 for the picture description, reverse counting and diadochokinetic rate tasks, respectively. When adding feature selection and semantic features, the accuracy rates increase to 1.00, 0.96 and 0.83 respectively; thus reaching very high classification accuracy on 3 different tasks.\n\nConclusionsWe show that speech-based features are highly predictive of medication state. Given that the highest performance was obtained with a very naturalistic task (picture description), our results suggest the feasibility of accurate, non-burdensome and high-frequency monitoring of medication effects.