Predicting Upper Limb Impairment in Stroke Patients Using Robotic Device Trajectory Images with Computer Vision

Learning Objectives:

• Describe how computer vision (CV) and machine learning can be applied to robotic device images to predict clinical motor impairment in stroke patients.
• Explain the process of generating and analyzing trajectory images from a KINARM exoskeleton to estimate Fugl-Meyer Assessment (FMA) scores.
• Discuss the potential clinical implications of using CV-based prediction models as adjunctive tools for stroke assessment and rehabilitation planning.