Micrographia in Parkinson's Disease: Automatic Recognition through Artificial Intelligence

Area of Science:

• Neurology
• Artificial Intelligence
• Biomedical Engineering

Background:

• Parkinson's disease (PD) is characterized by micrographia, a reduction in handwriting size.
• The early onset, progression, and L-Dopa responsiveness of micrographia in PD remain debated.

Purpose of the Study:

• To investigate the onset, progression, and L-Dopa responsiveness of micrographia in Parkinson's disease.
• To utilize artificial intelligence (AI)-based pen-and-paper handwriting analysis as a non-invasive tool.

Main Methods:

• Fifty-seven PD patients and 25 controls completed handwriting tasks.
• Handwriting samples were analyzed using both clinical (perceptual) and AI-based (automatic) methods.
• AI models, including convolutional neural networks (CNNs), were evaluated for classification accuracy.

Main Results:

• AI analysis confirmed micrographia in PD patients, manifesting early and worsening progressively.
• AI models accurately distinguished PD patients from controls (91%) and early from advanced PD (77%).
• Micrographia showed poor response to L-Dopa, and AI could not differentiate OFF/ON states.

Conclusions:

• AI-based handwriting analysis is a valuable, non-invasive tool for detecting and quantifying micrographia in PD.
• This AI approach supports telemedicine applications for Parkinson's disease management.
• AI analysis provides objective insights into PD progression and treatment response.