A BiLSTM-ENHANCED CNN ARCHITECTURE FOR EARLY DETECTION OF PARKINSON’S DISEASE

Abstract

A novel method for the detection of Parkinson's Disease (PD) through the analysis of spiral and wave drawings,
utilizing a hybrid Convolutional Neural Network (CNN) and Bidirectional Long Short Term Memory (BiLSTM) architecture.
PD diagnosis often involves motor function assessments, and drawing tasks present discernible patterns reflective of PDrelated motor impairments. Our proposed approach integrates a CNN for spatial feature extraction from the drawings and an
BiLSTM for sequential stroke analysis, enabling a comprehensive examination of motor patterns indicative of PD. This noninvasive and objective method holds promise for early-stage PD detection, offering quantifiable and accessible diagnostic
insights that could potentially facilitate timely interventions and improve patient outcomes. Automated analysis of drawingbased tasks through deep learning techniques represents an innovative avenue for PD