Superposition and modulation of muscle synergies for 1 reaching in response to a change in target location

23 We have recently shown that the muscle patterns for reaching are well described by the 24 combination of a few time-varying muscle synergies supporting the notion of a modular 25 architecture for arm control. Here we investigated whether the muscle patterns for reaching 26 movements involving on-line corrections are also generated by the combination of the same set 27 of time-varying muscle synergies used for point-to-point movements. We recorded end-point 28 kinematics and EMGs from up to 16 arm muscles of five subjects reaching from a central 29 location to eight peripheral targets in the frontal plane, from each peripheral target to one of the 30 two adjacent targets, and from the central location initially to one peripheral target and, after a 31 delay of either 50, 150, or 250 ms from the go signal, to one of the two adjacent targets. Time32 varying muscle synergies were extracted from the averaged, phasic, normalized EMGs of point33 to-point movements and fit to the patterns of target change movements using an iterative 34 optimization algorithm. In all subjects, three time-varying muscle synergies explained a large 35 fraction of the data variation of point-to-point movements. The superposition and modulation of 36 the same three synergies reconstructed the muscle patterns for target change movements better 37 than the superposition and modulation of the corresponding point-to-point muscle patterns, 38 appropriately aligned. While at the kinematic level the corrective trajectory for reaching during a 39 change in target location can be obtained by the delayed superposition of the trajectory from the 40 initial to the final target, at the muscle level the underlying phasic muscle patterns are captured 41 by the amplitude and timing modulation of the same time-varying muscle synergies recruited for 42 point-to-point movements. These results suggest that a common modular architecture is used for 43 the control of unperturbed arm movement and for their visually guided on-line corrections. 44