Parkinson's Disease (PD) is a neural disorder that can affect your ability to regulate movements. The disease usually starts slowly at early stages and gets worse as time passes. If you have Parkinson's Disease, it will become difficult to maintain your body balance and coordination. As the time passes the disease gets worse and may cause trouble in talking, sleeping, having mental and memory problems and many other related symptoms. This paper proposes two deep learning models for detecting PD. The first model proposed is a Convolutional Neural Network (CNN) model trained on hand drawn images. The second model is a Long Short Term Memory (LSTM) model trained using time series signals of hand drawings. The proposed models are trained using the benchmark NewHandPD dataset which consists of hand drawn images and its corresponding time series signals of both the healthy and PD subjects.
