Paper Title
A NOVEL APPROACH USING DEEP LEARNING FOR PARKINSON'S DISEASE DETECTION

Abstract
Parkinson's disease (PD) is an advancing movement ailment that has an impact on the nervous system.As of 2024, it is the second highest prevalent neurodegenerative illness and is rapidly increasing in prevalence.Individuals with Parkinson's may experience challenges in walking, speaking, and performing simple tasks as their symptoms advance. Over time, these symptoms can significantly impact well-being and quality of life. Analyzing image data from neuroimaging studies is made more effective along the utilization of Convolutional Neural Networks (CNNs), which provides deeper insights into brain changes linked to Parkinson's.The purpose of this project is to determine the efficacy of deep learning and machine learning approaches models in order to pinpoint the most exact method for detecting Parkinson's disease.We introduce a novel approach for PD detection utilizing deep learning methods. Our model incorporates a deep neural network that employs U-Net for image segmentation—generating a mask of the input image—and CNN for classification, identifying the class of the input.Our model achieved an overall accuracy of 98.98%, demonstrating its capability to detect PD with high precision. Keywords - Machine Learning, Deep Learning, U-Net, Healthcare