NeuroSynth: Unveiling the Parkinson's Palette - A Unified Ensemble Approach for Enhanced Predictive Modeling in Spiral Drawing Analysis

Abstract

Parkinson Disease (PD) is a progressive neurodegenerative illness, as it is marked by motor and non-motor deficiencies, and thus presupposing early and correct diagnosis to clinically manage it successfully. The present work suggests the machine learning and deep learning framework to be used in the initial prediction of PD by means of patient voice datasets. The proposed approach combines the process of preprocessing and feature extraction with such models as Support Vector Machine (SVM), Random Forests (RF), and Convolutional Neural Networks (CNN) and subsequently evaluates the performance in terms of such metrics as accuracy, precision, recall, and F1-score. As the experimental findings prove, the proposed method performs better in classification than the existing ones, which means that it can become one of the useful diagnostic instruments of clinicians. As noted in results, the application of artificial intelligence can help to improve predictability, timely diagnose, and provide personalized treatment plans to resolve Parkinson Disease.