A machine-learning approach to predict musculoskeletal load and running-related injuries using data from wearables

Running is a popular physical activity that entails, besides health benefits, also a risk of developing a running-related injury (RRI). As RRIs have a negative socio-economic impact, their prevention is important. Prediction of RRI occurrence is of interest as a potential vital factor in its prevention since it allows taking timely appropriate actions to avoid the development of an RRI. In addition, knowledge of structure-specific load (SSL) has the potential to improve the prevention or recovery of RRI since cumulative SSL has a pivotal role in RRI development. However, measuring SSL remains an unsolved challenge in applied settings. The combination of biomechanics, wearable sensors, and machine learning (ML) will be leveraged to estimate SSL and external load outside the laboratory. This load-estimation tool will be used to provide input to an improved RRI-prediction machine-learning model.