Machine learning-based detection of Parkinson's disease from facial expressions, hand movements, speech, and gait with general-purpose equipment: a systematic review.
Int J Med InformJul 5, 2026 (epub)
Clinical ResearchNeurology
Anastasiia A Moshkova, Anastasia O Novikova, Maksim I Polguev et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review evaluated machine learning (ML) methods for detecting Parkinson's disease (PD) using facial expressions, hand movements, speech, and gait with low-cost equipment. Analyzing 133 studies, it found that speech and voice modalities showed promising trends for classification. Multimodal data fusion improved performance in 75% of studies, although statistical validation was limited.
Key findings
- 74.5% of studies focused on binary classification tasks for PD detection.
- Support Vector Machines (SVM) and Random Forests (RF) were the most commonly used classifiers.
- 75% of studies indicated improved performance with multimodal data fusion.
Why it matters
The integration of machine learning with accessible technology could enhance the objectivity and scalability of Parkinson's disease screening and monitoring, potentially leading to earlier diagnosis and better patient management.