Using Machine Learning and Virtual Reality for Orthopedic Treatment and Abnormality Detection Based on Multivariate Time Series Data

In this work we present a virtual reality machine-learning system for telehealth orthopedic treatment. Our system can recognize orthopedic abnormalities and the presence of pain. It is based on a widely used virtual reality system, combined with its sensors. We implemented an algorithm that can identify very accurately wrist and neck pain and can serve as a real-time remote system for rehabilitation doctors or physical therapists, as part of a virtual reality telehealth treatment program. Our algorithms synchronize the patient’s movement data with a dedicated data server. The system has an easy-to-use interface for analysis of the collected data. We achieved more than 90% success rates evaluating the presence of neck pain and wrist pain across given exercises for each of our volunteers. Our system can serve as the basis for a real-world telehealth, clinically operative machine.