Visual feedback framework for rehabilitation of stroke patients

Abstract

The article describes an alternative for the rehabilitation of upper and lower limbs using visual feedback to treat the sequelae of stroke. The Framework contains three-dimensional scenarios which require patient immersion in an intuitive exercise routine supervised by the attending therapist, who further corroborates visually the progress of the recovery. This work is also supported by a DTW algorithm. The algorithm is based on the collection of data through tracking sensors and the identification of the exercise patterns performed by the patient. The delivering of a satisfactory response in case the exercise is completed correctly will support the evaluation of the therapist in the advance of the therapy. The experimental results show the execution of the exercises proposed with different users in order to validate the usability of the framework. Keywords: Stroke rehabilitation, Kinect v2, DTW algorithm, Unity, Matlab