Wearable sensors and machine learning for field-based biomechanical load assessment in sports: a review

The field-based assessment and management of biomechanical load, such asjoint forces, in sports has high significance f or eff ective athletic training andprevention of injuries. Developments in wearable sensors and machine learn-ing have improved assessment methods, as they enable measurements inreal training or competition environments outside the laboratory. This reviewsystematically summarizes the current state o f research on wearable sensorscombined with machine learning for assessing biomechanical load in sports.Searches were c onducted in PubMed and SPORTDiscus. A total of 5,426 arti-cles were identified, of which 42 met the eligibility criteria after screening.Data were e xtracted on participant characteristics, sports and movementtasks, wearable sensors used, machine learning methods, model input andbiomechanical output, validation strategies, and key findings. Running wasthe most frequently studied sport, although nine other sports w ere alsoinvestigated. Artificial neural networks and linear regression were the mostcommonly applied machine learning methods. Biomechanical load was mostfrequently assessed using ground reaction force metrics, followed b y mo ve-ment execution metrics and joint moment metrics. Current studies highlightthat wearable sensors and machine learning can predict biomechanical loads,especially in the lower e xtremities during running. Expanding the diversity ofsports studied and improving sensor placement on the upper extremities arenecessary to broaden the application of w earable sensors and machine learn-ing for athlete monitoring and injury prevention in real-world settings.