1Ph.D. Student, CIPCE, Human Motor Control and Computational Neuroscience Laboratory, School of ECE, College of Engineering, University of Tehran, Tehran, Iran.
2Associate Professor, CIPCE, Human Motor Control and Computational Neuroscience Laboratory, School of ECE, College of Engineering, University of Tehran, Tehran, Iran.
Central nervous system (CNS) uses an abundant set of joints and muscles to ensure both flexible and stable movements while interacting the environment. How the CNS faces the complexity of control problem and solves the question of physiological and mechanical abundances is not still clear. Modular control is one of the most prevalent hypotheses in answer to these questions. According to this point of view, CNS combines a few building blocks, here this will be muscle activities, named as muscle synergies, to present a vast repertoires of movements. In this study the algorithm of sample-based nonnegative matrix tri-factorization (NM3F) is used to extract spatial and temporal muscle synergy modules from muscle EMG data for three different types of point to point reaching (simple straight, reversal and via-point) movement in the frontal and sagittal planes. After extracting different features of the muscle synergies, physiological interpretation of these decomposed parts has been discussed. The first temporal module coded the direction and type of movement, while the spatial modules describe some via postures. Also the extracted modules are not similar for subjects. The recruitment of the spatial and temporal modules are correlated due to the movement direction.