Is there a reliable and invariant set of muscle synergy during isometric biaxial trunk exertion in the sagittal and transverse planes by healthy subjects

Abstract It has been suggested that the central nervous system simplifies muscle control through basic units, called synergies. In this study, we have developed a novel target-matching protocol and used non-negative matrix factorization (NMF) technique to extract trunk muscle synergies and corresponding torque synergies. Isometric torque data at the L5/S1 level and electromyographic patterns of twelve abdominal and back muscles from twelve healthy participants (five females) were simultaneously recorded. Each participant performed a total number of 24 isometric target-matching tasks using 12 different angular directions and 2 levels of uniaxial and biaxial exertions. Within- and between-subject similarities were assessed by considering both the data of different pairs of participants, where the activation coefficients of one participant were used in the NMF analysis of another participant, and the Pearson’s correlation coefficients ( R ) between muscle synergy vectors. The results showed that, for a healthy person, a set of four muscles (overall variance accounted for (VAF) of 97.9±0.53%) and four corresponding torque synergies (overall VAF of 92.2±3.03%) could efficiently decompose the sagittal and transverse torque planes into their main directions. Furthermore, the correlation coefficients were 0.77±0.12, 0.86±0.08, 0.78±0.12, and 0.93±0.04, for all synergies, reflecting the consistency of muscle synergies across participants. Overall, our results suggest that by taking advantage of muscle synergies we could potentially overcome the redundancy inherent to control strategies of the trunk neuromuscular system. In future studies, the synergies identified in patients with low back pain could be compared with those extracted from healthy participants towards various clinical and rehabilitation applications.