A Kinect-Based System for Stroke Rehabilitation

Abstract

Virtual reality (VR)-based stroke rehabilitation has been shown to be effective in increasing motivation and functional performance in stroke patients. The new motion-sensing technology, Kinect, is cost effective and does not require the patient to wear sensors on the body, which increases freedom of movement. The objective of this study was to use Kinect technology to develop a VR stroke rehabilitation system with unilateral and bilateral tasks for recovering the function of the upper extremity. This study tested the feasibility, therapeutic effectiveness, and user acceptance of this technology. Two participants with various levels of motor severity received 30-minute stroke rehabilitation 3 times per week over 8 weeks (a total 24 training sessions). The Wolf Motor Function Test (WMFT), Test Évaluant la performance des Membres supérieurs des Personnes Âgées (TEMPA), and Fugl-Meyer Assessment of Physical Performance (FMA) were used to collect data before and after rehabilitation, and during a follow-up to detect the changes of functional performance. Questionnaires of user acceptance of the technology were administered. On completion of the rehabilitation program, using the proposed Kinect-based VR training system, WMFT, TEMPA, and FMA results increased for both participants. The technology acceptance questionnaires indicated that participants had strong intentions to continue using the proposed system for rehabilitation. We developed the first Kinect-based stroke rehabilitation for the upper extremity, and demonstrated its feasibility and effectiveness in improving upper extremity function after a stroke. A large-scale study should be conducted to test the effectiveness of the proposed system for stroke rehabilitation.