Drawing Patterns Could Signal Parkinson's Early

A study using drawing tests may detect Parkinson's disease with high accuracy, offering potential for earlier intervention.

Siksha 'O' Anusandhan University researchers published early results in *Discover Computing* (2026) showing an AI system could identify Parkinson's disease through hand movement patterns in drawing tasks. The analysis used biometric smart pens to capture participants’ motions during meander and spiral drawing exercises. Among 66 people—31 with Parkinson’s and 35 healthy controls—the AI named SNAKE achieved 98.95% accuracy for meander patterns and 97.74% for spirals. This non-invasive approach could enable earlier detection, potentially reducing healthcare costs and improving quality of life through timely intervention.

The method works by translating subtle hand tremors and movement irregularities into digital data. AI then identifies deviations from typical patterns that correlate with Parkinson’s progression. Unlike blood tests or scans, this technique requires no physical contact or radiation, making it accessible in community health settings.

For health access, early detection through low-cost, non-invasive screening could free up resources for advanced care while delaying disease progression. This matters most for low-resource regions where Parkinson’s diagnosis often comes too late. However, the current study’s small sample size may not reflect Parkinson’s diversity across global populations, and the results do not yet constitute clinical diagnostic proof.

Next steps include larger trials across diverse demographics to validate scalability. Until then, this framework shows promise for early health monitoring but remains a research tool—not a ready solution.

Source: ScienceAlert