Parkinson’s disease (PD) is a complex age-related progressive, degenerative disorder of the central nervous system. There is no definitive test for the diagnosis of PD; the disease must be diagnosed based on clinical criteria. Several studies have found that vocal impairment may also be one of the earliest symptoms for the onset of the PD. This finding has drawn attention to consider voice and speech in detecting and monitoring the progression of PD. Consequently, the main objective of this article is to investigate the effectiveness of several techniques in conducting the PD diagnostic problem. Aiming at improving the efficiency and effectiveness of the classification accuracy for PD diagnosis this article proposes a two-stage feature reduction technique. This paper purposes hybrid algorithms for addressing the fore mentioned problem with sufficient statistical inference testing and present experimental results with quantitative and qualitative analysis. The proposed inference system includes several highly efficient statistical learning algorithms for the evaluation of the voice features of the PD controls. Several traditional methods are used for the measurements of the voice features based on fundamental voice frequency oscillation, jitter measure, shimmer test, pitch period entropy noise to harmonics ratios. These measures are used to examine the extent of the Parkinson disease in clients.
Volume 11 | 08-Special Issue
Pages: 2162-2174