Transforming Parkinson’s Disease Diagnosis Using Deep Neural Network

Parkinson’s disease diagnosis is a critical medical challenge, and machine learning (ML) presents a promising avenue for early detection. This study evaluates the performance of six ML classifiers SVC, DecisionTreeClassifier, AdaBoostClassifier, RandomForestClassifier, KNeighborsClassifier, XGBClassifier, and deep learning model DNN on two datasets, MDVP, and Oxford Parkisons Disease. In MDVP, DNN leads with 97.21% accuracy, outperforming others. On the Oxford dataset, Deep Neural Network (DNN) again excels with 94.11% accuracy. These results highlight DNN’s consistent superiority in PD detection across datasets. This study contributes towards the enhancement of the use of the ML tool for the diagnosis of PD for greater efficiency in early intervention. Further investigation of ML and deep learning in this domain has a great potential to change the status quo treadmill. The nature and functioning of diagnostic activities in an unfolding healthcare system.