Selection of optimal parameters for automatic analysis of speech disorders in Parkinson's disease

Abstract

Patients with Parkinson's disease (PD) usually suffer from hypokinetic dysarthria (HD), which involves impairment of phonation, articulation, prosody, and speech fluency. Our paper deals with parameters that can be used for the evaluation of motor aspects of speech and relevant methods of data acquisition and analysis. A review of specific parameters of HD and methods used for their evaluation may from the practical point of view contribute both to the diagnostic approaches to HD and to the development of suitable measures for assessment of its progression. The paper gives a description of the most frequently used parameters and their optimization to enable the best possible automatic classification of the various stages of Parkinson's disease.