Diagnosis of Parkinson’s Disease by Deep Learning Techniques Using Handwriting Dataset

Diagnosis and evaluation of Parkinson’s disease (PD) by clinicians is normally dependent on several established clinical criteria. Measuring the severity level according to these criteria depends heavily on the doctor’s expertise, which is subjective and inefficient. Hence, the current work aims to provide a quantitative and comprehensive evaluation in order to improve the diagnosis and assess the severity of PD using deep learning approaches. In the current study, a machine learning based method is proposed to automatically diagnose the PD severity from handwritten exams, by extracting the information from static and dynamic tests. A hybrid model for prediction is suggested by fusing convolutional neural network (CNN) and long-short term memory (LSTM). The suggested hybrid model (CNN + LSTM) obtained accuracies of 85.5% and 99.3% using the ParkinsonHW and HandPD datasets, respectively. We conclude that the quantitative evaluation provided by our model may be considered a helpful tool in the clinical detection of Parkinson’s disease.