Early Detection of Parkinson's Disease using Machine Learning & Image Processing
M. Nithya, V. Lalitha, K Paveethra, S Kumari
Abstract
Parkinson Disease (PD) is a brain disorder which affects the central nervous system such as shaking, stiffness, and difficulty with walking, balance, and coordination. Since PD is closely associated to other neurological symptoms, it is generally difficult to accurately predict the disease. Further the close association of PD symptoms with other neurological symptoms results in more than 25% of wrong detection of PD. Therefore, the proposed system focuses on developing an automated diagnosis system based on Machine learning (ML) which can exactly predict the PD & healthy control (HC). Weighted Magnetic Resonance Imaging (MRI) for PD and HC are provided by Parkinson's Progression Markers Initiative (PPMI). Image registration technique is used to align midbrain slices. Damaged brain pixel is detected using hybrid technique (SVM and Random forest) algorithm. The results conclude Machine Learning (ML) offers better accuracy and specificity.