LOW COST, ADAPTABLE, AND CLINICIAN-FRIENDLY VIRTUAL REHABILITATION SYSTEM USING THE MICROSOFT KINECT

Abstract

Virtual rehabilitation (VR) provides several advantages over conventional therapy. These include an increased capacity to be developed at low cost by exploiting commercially available gaming technology, deliver real-time performance feedback through visual and auditory modalities, obtain quantitative measures of progress, enhance motivation and thereby improve adherence, and, importantly, have the potential to deliver patient-specific treatment that adapts with functional improvements over training. Widespread adoption will enable frequent data gathering and analysis that will enable physical therapy to advance through evidence of patient progress. Successful applications could dramatically alter physical therapy practice and optimize rehabilitation outcomes. Despite these significant benefits, clinical adoption of VR is slow. One reason is that current systems are typically constrained to providing therapy for one specific motor deficit with a limited number of pre-defined therapeutic game exercises. Any adjustments to these exercises, patient feedback, and difficulty levels require high level programming knowledge, which is typically beyond the therapistsÍ skill level and time afforded to treat patients. With our adaptable MATLAB-based VR system using the Microsoft Kinect, a clinician can easily develop therapeutic game exercises as well as assess patient functionality. When the patient uses the system at home, the collected data will ensure patient compliance with the prescribed home therapy. At the same time, analysis of the data will provide information about the efficacy of the treatment. By overcoming the limitations of current VR systems, our system is preparing the ground for wider adoption of such systems in the future.