Human activities are documented from the simplification of the redundant degree of freedom

Abstract

Human achieves complex and redundant movements, such as standing and walking. Central nervous system (CNS) coordinates huge degree of freedom of the musculoskeletal system. To this end, muscle activities were accounted for with low-dimensional sets of muscle synergies. Present review focused on the muscle synergies to comprehend human movements from the simplification of the redundant degree of freedom. The muscle synergies during human movements were extracted from the data matrix of recorded EMGs of lower limb muscles using non-negative matrix factorization. During postural maintenance in horizontal plane, the CNS flexibly changed patterns of the muscle synergy recruitment to achieve effectively postural control. During gait transition between walking and running, the muscle synergies and their activation profiles dramatically changed when a gait transition was observed. To examine the functional role for existing muscle synergy, learning speed of neural network model with and without muscle synergy layer was calculated. As a result, the learning speed was higher with muscle synergy layer than without it, because the muscle synergy reduces the bias in the mechanical direction of the muscles. From these investigations, present review concluded that complex and redundant human activities are enabled under existence of low-dimensional sets of muscle synergies.