Automated Upper Extremity Rehabilitation for Stroke Patients Using a Partially Observable Markov Decision Process

Patricia Kan, Jesse Hoey, Alex Mihailidis

This paper presents a real-time system that guides stroke patients during upper extremity rehabilitation. The system automatically modifies exercise parameters to account for the specific needs and abilities of different individuals. We describe a partially observable Markov decision process (POMDP) model of a rehabilitation exercise that can capture this form of customization. The system will be evaluated in user trials during summer 2008 in Toronto, Canada.

Submitted: Sep 11, 2008