Prediction of Parkinson’s Disease Based on Spiral and Wave Drawings Using Deep Learning and Explainable AI

Authors: Akshay Lamkhade, Prasad Khairnar, Smaksh Gupta, Narendra Wakhare, Sandesh Deshmukh

DOI: https://doi.org/10.37082/IJIRMPS.v13.i3.232440

Short DOI: https://doi.org/g9g7j7

Country: India

Full-text Research PDF File:   View   |   Download

Abstract: Parkinson's Disease (PD) is a progressive neurodegenerative disorder that affects motor skills and cognitive functioning. Accurate and early diagnosis of PD can significantly impact treatment outcomes and quality of life for patients. This paper presents a novel approach for early PD prediction based on the analysis of hand-drawn spiral and wave patterns using Explainable Artificial Intelligence (EXAI) techniques. By integrating convolutional neural networks (CNNs) such as VGG19 and GoogleNet with the LIME (Local Interpretable Model-Agnostic Explanations) framework, this system ensures both high accuracy and transparency in predictions. Our proposed model achieves an accuracy of 98.45%, outperforming traditional black-box approaches. The integration of EXAI enhances the interpretability of the model, making it suitable for clinical application and improving clinician trust.

Keywords: Parkinson’s Disease, Explainable Artificial Intelligence, Convolutional Neural Networks, Spiral and Wave Drawings, VGG19, GoogleNet, LIME

Paper Id: 232440

Published On: 2025-05-01

Published In: Volume 13, Issue 3, May-June 2025

Share this