The study proposes the use of machine learning techniques and Inertial Measurement Units (IMUs) to monitor and evaluate the execution of prescribed physical therapy exercises at home, particularly for the elderly. The system can identify the exercise being carried out and assess its correctness based on data from four IMUs placed on the person’s limbs. The study demonstrates the potential of IMUs in providing a complete monitoring solution for physical therapies, which could be utilized in the design of virtual coaches.
Publication date: 25 Jan 2024
Project Page: https://zenodo.org/record/50527
Paper: https://arxiv.org/pdf/2401.12857
