Enhancing Cross-Patient Generalization in AI-Based Parkinson s Disease Detection

TL;DR

This paper introduces a two-stage AI approach for Parkinson's disease detection from hand-drawn images that improves cross-patient generalization by dividing images into chunks and using ensemble classification, outperforming state-of-the-art methods.

Contribution

The paper proposes a novel chunking and ensemble strategy that enhances robustness and accuracy in Parkinson's detection, especially on unseen patients.

Findings

Achieved 97.08% accuracy on seen patients

Achieved 94.91% accuracy on unseen patients

Reduced performance gap between seen and unseen patients to 2.17%

Abstract

Parkinson's disease (PD) is a neurodegenerative disease affecting about 1% of people over the age of 60, causing motor impairments that impede hand coordination activities such as writing and drawing. Many approaches have tried to support early detection of Parkinson's disease based on hand-drawn images; however, we identified two major limitations in the related works: (1) the lack of sufficient datasets, (2) the robustness when dealing with unseen patient data. In this paper, we propose a new approach to detect Parkinson's disease that consists of two stages: The first stage classifies based on their drawing type(circle, meander, spiral), and the second stage extracts the required features from the images and detects Parkinson's disease. We overcame the previous two limitations by applying a chunking strategy where we divide each image into 2x2 chunks. Each chunk is processed separately when extracting features and recognizing Parkinson's disease indicators. To make the final classification, an ensemble method is used to merge the decisions made from each chunk. Our evaluation shows that our proposed approach outperforms the top performing state-of-the-art approaches, in particular on unseen patients. On the NewHandPD dataset our approach, it achieved 97.08% accuracy for seen patients and 94.91% for unseen patients, our proposed approach maintained a gap of only 2.17 percentage points, compared to the 4.76-point drop observed in prior work.