Training Machine Learning Classifiers on Differentiating Major Depressive Disorder and Parkinson's Disease Using Cognitive Performance

Machine-learning (ML) algorithms are extensively employed to develop early detection tools for neurological and psychiatric diseases. Parkinson's disease (PD) and major depressive disorder (MDD) have significantly high comorbidity and prodromal presentation. Consequently, early detection can help improve the quality of life of MDD and PD patients. Here, we examined whether distinguish between patients with MDD and PD. We deployed five machine-learning classifiers to differentiate MDD from PD. The best test classification was obtained from logistic regression classifier with an accuracy of 91 %. To our knowledge, this is the first study to apply machine learning algorithms on cognitive performance to differentiate between patients with MDD and PD.