Motor control based on the muscle synergy hypothesis

Abstract

In neuroscience, the idea that motor behaviors are constructed by a combination of building blocks has been supported by a large amount of experimental evidences. The idea has been very attractive as a powerful strategy for solving the motor redundancy problem. While there are some candidates for motor primitives, such as submovements, oscillations, and mechanical impedances, synergies are one of the candidates for motor modules or composite units for motor control. Synergies are usually extracted by applying statistic techniques to explanatory variables, such as joint angles and electromyography (EMG) signals, and by decomposing these variables into fewer units. The results of factor decomposition are, however, not necessarily interpretable with these explanatory variables, even though the factors successfully reduce the dimensionality of movement; therefore, the physical meaning of synergies is unclear in most cases. To obtain insight into the meaning of synergies, this chapter proposes the agonist-antagonist musclepair (A-A) concept and uses other explanatory variables: the A-A ratio, which is related to the equilibrium point (EP), and the A-A sum, which is associated with mechanical stiffness. The A-A concept can be regarded as a form comparable to the EP hypothesis (EPH, λ model), and it can be extended to the novel concept of EP-based synergies. These explanatory variables enable us to identify muscle synergies from human EMG signals and to interpret the physical meaning of the extracted muscle synergies. This chapter introduces the EMG analysis in hand-force generation of a human upper limb and shows that the endpoint (hand) movement is governed by two muscle synergies for (1) radial movement generation and (2) angular movement generation in a polar coordinate system centered on the shoulder joint. On the basis of the analysis, a synergy-based framework of human motor control is hypothesized, and it can explain the mechanism of the movement control in a simple way.