Identification of shoulder muscle synergies in healthy subjects during an isometric task.

ABSTRACT

Muscle Synergy method has been proposed in the literature to provide a lower dimensional representation of motor commands from the central nervous system (CNS). Studies on post-stroke patients highlighted how features such as the minimum number of motor synergies accounting for most of the data variance correlate with impairments and motor function. In this study, we target healthy subjects to establish normative data in isometric tasks involving shoulder muscles. Five subjects performed an isometric, two-dimensional force-matching task in twelve planar directions with two force levels across shoulder joint. Muscle synergies and their respective activation curves were computed from nine upper limb muscles via a nonnegative matrix factorization (NNMF) algorithm. Four synergies, on an average, were able to explain 95% of the variance in EMG datasets across all subjects. The cosine similarity of the muscle synergies among the subjects on an average is found to be 0.79±0.20. Two subjects revealed the presence of subject-specific synergies which will require further investigation before examining impaired subjects.