Comparing estimated and measured muscle activation during highly dynamic and multidirectional movements: a validation study

The purpose of this study was to validate muscle activation of the lower extremities computed in AnyBody with measured muscle activations (EMG) in highly dynamic movement tasks. Ten participants performed walking, jogging, sprinting and cutting tasks. Kinetic, kinematic and EMG data were captured for 8 muscles of the dominant leg. The average correlation coefficient (CC) was 0.51 (max.: 0.83, min.: -0.05) with 71% of all trials showing moderate to very good compliance. The average mean absolute error (MAE) was 1.32 (max.: 3.71, min.: 0.17). Co-contraction, precision of the muscle recruitment algorithm, electromechanical delay and anthropometrical measures may have affected the results. The estimation of computed muscle activation can be a suitable method for certain muscles considering highly dynamic movement tasks.