The use of non-linear acoustic analysis to objectively evaluate the voice of people with Parkinson's disease

Voice acoustic analysis in people with Parkinson's disease can be a valuable and objective tool supporting the diagnosis of neurodegenerative diseases. The article analyses the application of non-linear coefficients for the parametrization of the speech signal representing the vowel “a” with prolonged phonation in patients with Parkinson's disease. The study was conducted using three different non-linear techniques for the generation of speech signal features. The research utilized recordings conducted at the Department of Neurology at the Medical University of Warsaw. The discriminatory ability of individual coefficients was evaluated using the LDA technique and k-nn method. The recognition accuracy is about 85 %.