Analysis of Kinematic Features of Handwriting Data to Detect Parkinson's Disease

Across various neurodegenerative diseases, the second most common disease is Parkinson's disease. It affects the brain cells, which are accountable for releasing dopamine. When there is less release of dopamine, the muscles of the body can't work properly, which creates a problem for a person. In recent years, it caught the eyes of researchers that handwriting analysis can be a good solution to identify Parkinson's disease. Various attempts have already been made since the discovery of this idea. In this work, we have tried to use kinematic features of handwriting to detect Parkinson's disease. As there was less data variational autoencoder was used to generate synthetic data to train the model and the obtained accuracy of this proposed model is 89%