Assessing clinically relevant biomechanical biomarkers in the field to predict physical functioning and health in patients with knee osteoarthritis: a nation-wide citizen science approach

Wider research context: Over the last three decades, marker-based 3D motion capture systems have become paramount in accurately quantifying and analyzing human movement in clinical settings for both research and decision-making purposes. However, despite their value in clinical settings, human movement science and biomechanical research, their high costs, complex and laboratory-bound setups, and need for trained operators limit their wide-spread use and impact for our societies. Recent advancements in deep learning and computer vision have led to the development of 3D multi-camera (smartphone-based) markerless motion capturing, enabling the tracking of human locomotion from simple videos in the field. Objectives: This project aims to explore the potential of out-of-the-lab markerless-driven biomechanical assessments in predicting clinically relevant changes in knee osteoarthritis patients undergoing physiotherapy. By utilizing an open-source 3D markerless and machine-learning drive motion capture system based on two simple smartphones (www.openCap.ai), the study will assess 3D movement kinematics and kinetics in real-world settings and correlate these biomechanical parameters with patient characteristics and self-reported health outcomes. Methods: 25 Physiotherapists of the Austrian GLA:D network, will serve as our Citizen Scientists, and will evaluate patients with knee osteoarthritis using established patient reported outcomes and clinical examination, along with capturing movement dynamics during walking, sit-to-stand, and squatting using the open-source OpenCap.ai markerless motion capture system. A highly diverse sample of approximately 165+ patients will be included over 1.5 years, and data will be assessed at baseline, after a six-week rehabilitation program, and at 12-month follow-up. Correlations will be used to analyze associations between biomechanical parameters, patient characteristics, and health outcomes.