A Virtual Haptic Motor Rehabilitation System

Abstract

There are over 40,000 strokes in Canada each year, with a 75% rate of physical impairment. Virtual Rehabilitation was proposed as an alternative for recovering stroke patients, since conventional rehabilitative methods limit patients with time constraints. Our research presents a virtual rehabilitation system, consisting of a series of applications and using the Immersion CyberForceRTM system. The system is designed to improve patients' abilities to perform everyday tasks, such as drinking tea, as per the feedback of occupational therapists. It also includes an interface for occupational therapists to configure each exercise. We designed a progress measurement component for automated advancements without requiring an occupational therapist to analyze performance results. An LSR Analyzer normalizes the raw sensor values read from the Immersion CyberGlove, for better sensor readings to be fed into an Artificial Neural Network. This is used to minimize the requirements of the calibration process. Analysis of all results and feedback is presented.