Freezing of Gait Prediction in Parkinsons Patients Using Neural Network

Parkinson's disease is a neural disease that slowly affects the nerves of a person ultimately leading to death. In the later stages of the disease, patients tend to experience a phenomenon called freezing of gait, which basically means utter incapacitation of the patients' limbs. The main objective of freezing of gait prediction in Parkinson's patients using the neural network is to predict, in advance, if the limbs of a Parkinson's Patient are going to freeze. This is to send them the required assistance at the right time, without any delay, to avoid any injuries that might occur to the patients concerned. We implement Neural Networks to perform this prediction. Specifically, a multilayer perceptron is implemented to provide results. This multilayer perceptron is implemented in R language, which is a programming language for statistical computing processes. The proposed system, FGPuMLP, achieves lesser error compared to the logistic regression model.