A Stroke of Insight: Accessible Screening for Cervical Myelopathy via Tablet and Paper Drawings

Cervical myelopathy (CM) is a progressive spinal disorder affecting hand function. We developed two machine learning-based screening methods: one using tablet drawings and another using smartphone-captured paper drawings. In the tablet method, users trace spiral, square, and triangular figures with a stylus, which a convolutional neural network classifies with 95% sensitivity and 85% specificity, exceeding conventional examinations. The smartphone method captures paper drawings, enabling screening without specialized equipment. Together, these approaches provide accessible and reliable CM screening in clinical and non-clinical settings.