Multimedia non-invasive hand therapy monitoring system

In this paper, we present a novel non-invasive multimedia therapy retrieval framework to assist therapists in evaluating and managing the progress of Hemiplegic patients using inexpensive off-the-shelf 3D motion sensor. We have developed therapy retrieval analytics to detect, recognize and track key rotational and angular movements of different hand joints. The framework stores the raw motion data in a network repository, retrieves the raw motion data on demand and produces live plots that show quality of improvement metrics such as range of motion of each joint movement in the hand. These statistical reports would assist therapists and caregivers to get an insight about the progress of a patient´s therapy and thus help adapting her/his future therapy assignments accordingly. Finally, we will present the analysis of initial test results that show the viability and potential of our system.