Parametric analysis of pilot voice signals in Parkinson?s disease diagnostics

Parkinson's disease (PD) is a neurodegenerative of the central nervous system (CNS) characterized by progressive loss dopaminergic neurons in substantia nigra. The article describes an analysis pilot voice signal diagnostics. Frequency domain was mainly used to assess state patient's apparatus order support PD recordings covered uttering ?a? sound at least twice with extended phonation. research utilized real acquired Department Neurology Medical University Warsaw, Poland. Spectral speech coefficients may be determined based on different defined frequency scales. authors four scales: linear, Mel, Bark and ERB . descriptors have been for each scales which are widely machine deep learning applications, perceptual analysis. usefulness extracted features assessed taking various methods into account. discriminatory ability individual evaluated using Fisher coefficient LDA technique.. results numerical experiments shown efficiencies proposed frequencies