Metric-Based Muscle Synergy Consistency for Upper Limb Motor Functions | IEEE Journals & Magazine | IEEE Xplore

Metric-Based Muscle Synergy Consistency for Upper Limb Motor Functions


Abstract:

The hypothesis of muscle synergy has been dominantly accepted in research due to the evidence of being well investigated its existence in multiple disciplines, such as he...Show More

Abstract:

The hypothesis of muscle synergy has been dominantly accepted in research due to the evidence of being well investigated its existence in multiple disciplines, such as health science. Motor function assessment using muscle synergy provides a quantitative approach from the physiological features of muscles. Given the synergy coordinating multiple muscles in a natural motion, this article proposes a surface electromyography (sEMG)-driven metric to measure the consistency of the muscle synergy patterns of upper limb motor functions. The metric-based consistency algorithm primarily comprises three stages: 1) to construct nonnegative matrix factorization (NNMF)-based sEMG muscle synergy patterns; 2) to validate synergy similarities via cosine similarity; and 3) to conduct the patterns consistency between motion cycles checking via the Mahalanobis distance (MD). The proposed measure is employed to validate synergy consistency in upper limb motor function of isokinetic movement protocols involving ten subjects. The muscle synergy similarity has achieved averages of 88% and 97% for intersubject and intrasubject, respectively. The synergy consistency via MD is significantly positively correlated with the maximum joint torque fluctuation between motion cycles to indicate its efficiency. This quantitative measurement confirms that muscle synergy can be used to measure muscle motor function, paving the way for motion protocol assessment, such as stroke rehabilitation and exercise-induced muscle injury avoidance through monitoring the consistency of the linear combination of muscle activations during movement.
Article Sequence Number: 4002411
Date of Publication: 03 December 2021

ISSN Information:

Funding Agency:


References

References is not available for this document.