Telerehabilitation Outcome Prediction via Machine Learning

In this study, we propose a telerehabilitation system designed to remotely monitor patients and analyze data collected from commercial wearable devices throughout a 12-weeks physical exercise program. The system, developed under the RAPIDO project, utilizes a Random Forest (RF) model to predict patients' rehabilitation outcomes based on health metrics collected during the program. Preliminary results show that the RF model can effectively predict rehabilitation outcomes, offering valuable insights into patient progress and supporting the personalization of rehabilitation programs by only leveraging on commercial devices.