Video-based quantification of patient's compliance during post-stroke virtual reality rehabilitation

摘要

We present a video-based monitoring system for quantification of patient's attention to visual feedback during robot assisted gait rehabilitation. Patient's face and facial features are detected online and used to estimate the approximate gaze direction. This gaze information is then used to calculate various metrics of patient's attention. Results demonstrate that such unobtrusive videobased gaze tracking is feasible and that it can be used to support assessment of patient's compliance with the rehabilitation therapy.