Assessing the Influence of Robot-Assisted Movement Through Muscle Synergies Similarities

There is a high interest in using assistive robotic devices to help people recover from neurological impairments and promote motor re-learning. These robots enhance users' overall motor control by facilitating the use of muscle synergies similar to those observed in healthy individuals. However, muscle synergies have yet to be used in the design of control strategies for rehabilitation devices. In this work, we propose a novel synergy-based objective function that assesses similarities in muscle synergies and describes how changes in robotic assistance levels impact muscle synergies. We asked healthy participants to execute a 2D reaching task with a collaborative robot in the transparent mode and then with three levels of assistance while holding a weight. Seven muscles' EMG signals were decomposed into muscle synergies for all the conditions. First, a pool of healthy synergies is presented as reference synergies. Then, we compared the reference synergies with those extracted when participants performed the task holding a weight with multiple assistance levels. The results showed slight variations in spatial muscle synergies and significant changes in the temporal synergies. Finally, we introduced a novel objective function that assesses participants' motor control and coordination when reaching a point with added weight. This preliminary stage shows promising results but is a first step towards improving the control of robotic rehabilitation devices.