Segmentation and observation of hand rehabilitation exercises by supporting of acceleration signals

In recent years, there has been growing interest in using sensors worn on patients' bodies to assist doctors in monitoring and evaluating rehabilitation therapy. In this study, we toward an assisted tool that aids in the rehabilitation assessments of patients with impaired hands. The proposed method automatically spots the patient's exercises using utilizing accelerometers and displays the results on the collected video. This tool helps doctors to monitor and evaluate the patient's progress more concentrated and easily. The proposed method utilizes wavelet signal transformations, and signal segmentation by Sliding Window and Short-time Energy techniques. These techniques are not only simple calculations but also offer promising results. In the experimental results, the exercises, which are set up in real clinical situations, have been segmented with 87% of accuracy and 80% of sensitivity rate. These results confirm a potential solution to improve the efficiency and effectiveness of the rehabilitation process using the automatic assisted application.