Muscle-Synergy for Human-Robot Control, Wearable Devices and Computer Animation: A Survey

Abstract

Known studies in neurosciences have discovered that locomotion or motor tasks in humans involve the central nervous system (CNS) to manage activation signals for several groups of muscles. The CNS produces coordinated activation patterns, known as muscle synergy, that reduce the number of control signals by combining them into sets of activation signals. Muscle synergy is possible to be observed via electromyography (EMG) and can be extracted from the EMG data through factorization. As several factorization methods can be used to extract muscle synergies, non-negative matrix factorization (NMF) was suggested best for active tasks. While the concept of muscle synergy has been adapted in various disciplines such as healthcare, engineering and computer graphics, this narrative review discusses how muscle synergy has benefited them. Application involving muscle synergies is presented that focuses on muscular rehabilitation, human-robot control, prostheses assistive devices and graphical animation. Finally, future research is conjectured about technical challenges and prospects of muscle synergy.