Early Prediction of Parkinson Disease using Deep Learning

Abstract

Parkinson’s Disease (PD) is one of the most critical progressive neurological diseases mainly affect the motor system of the body. The accurate diagnosis of PD has been a challenge, mainly due to the close relevance of PD with other neurological diseases. This close relevance causes 25% inaccurate manual diagnosis of PD. Many researchers have suggested different algorithms and techniques for predicting Parkinson Disease. But after the overall analysis, Deep Learning Algorithm offers superior detection performance. Convolutional Neural Network (CNN) is used to classify PD and Healthy Control (HC) patients accurately. Parkinson Progression Markers Initiative (PPMI) provides publically available benchmarked Magnetic Resonance Images (MRI) images for both PD and Healthy Control (HC). These images are used to train the model. Proper tuning of parameters in CNN helps to reduce error rates, thus making the model more reliable.