Optimized Q-Learning based Fuzzy System for Early Diagnosis and Assessment of Parkinson’s Disease

The emerging disease in today’s era is the Parkinson ’s disease (PD) is a mainly affects the motor and cognitive functionalities and is a progressive neurodegenerative disorder for humans those are above age 60. The early and exact analysis is essential for successful intervention. The older diagnostic methods are cost-efficient and time-consuming in order to do the clinical evaluations and imaging techniques. The proposed method implements the Q-learning with fuzzy logic for early diagnosis for identifying the impact of PD. Furthermore, the fuzzy membership functions and reward mechanisms, integrating thereby improving adaptive learning improving the prediction accuracy. The proposed model is trained and tested using the fuzzy membership function and reward function incorporating key clinical parameters such as hand tremors, voice impairment, cognitive decline and gait instability, including motor and non-motor skills for early diagnosis. The fuzzy inference system classifies the severity of the disease into multiple levels, while the Q-learning component continuously refines decision-making based on real-time feedback. The proposed model’s ability to provide a reasonably accurate estimation of disease severity is 91%.