TherapEase: Enhancing Remote Physical Therapy Using Pose Estimation, Emotion Recognition, and K-Nearest Neighbors (KNN) Algorithm

The COVID-19 pandemic has accelerated the adoption of telerehabilitation, revealing both opportunities and challenges in remote physical therapy. This study introduces TherapEase, a web-based application designed to enhance the quality of virtual rehabilitation care by addressing issues such as exercise performance accuracy and communication clarity. TherapEase aims to complement, rather than replace, on-site physical therapy. A notable feature of the system is its implementation of pain assessment during therapy sessions through emotion recognition, utilizing the You Only Look Once (YOLO) algorithm to evaluate patients' facial expressions. Developed using an agile methodology, the application underwent evaluation based on the ISO 25010:2011 software quality model, focusing on Functional Suitability, Performance Efficiency, Interaction Capability, and Reliability. Results indicate high satisfaction levels across all evaluated aspects, with mean scores ranging from 3.58 to 3.67 out of 4. These findings suggest that TherapEase effectively integrates various components to provide a coherent and efficient telerehabilitation solution, potentially improving the delivery of remote physical therapy services while supporting traditional in-person care. Further research is needed to improve the accuracy of the pain assessment feature.