Explainable deep learning architecture for early diagnosis of Parkinson’s disease

Parkinson’s disease (PD) is a neurodegenerative disease that develops in middle-aged and older adults. The development of a gait detection for PD patients to assist doctors in diagnoses is a crucial research target. This work develops an explainable learning architecture that involves deep learning, machine learning, data selection, feature evaluation and data balancing mechanisms, for gait detection in PD patients. The results obtained for these patients and healthy individuals are analyzed with different algorithms. Application of the research architecture that is developed in this study significantly increased the overall rate of identification; specifically, when used proposed architecture with XGBoost and deep learning achieves accuracy rates as high as 97.32% and 98.41%, respectively. The goal is to provide an early detection system that can serve as a reference for doctors when making a diagnosis.