Affiliations: [a] Department of Biomedical Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran | [b] Medical Physics and Biomedical Engineering Department, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran | [c] Research Center of Biomedical Technology and Robotics, Tehran University of Medical Sciences, Tehran, Iran | [d] Department of Neurology, Mashhad University of Medical Sciences, Mashhad, Iran
Correspondence:
[*]
Corresponding author: Amir Homayoun Jafari, Medical Physics and Biomedical Engineering Department, School of Medicine, Tehran University of Medical Sciences, Tehran 1417613151, Iran. Tel.: +98 21 66466383, Fax: +98 21 66482654; E-mail:[email protected]
Abstract: OBJECTIVE: Muscle synergy is the activation of a group of muscles
that contribute to a particular movement. The goal of the present study is
to examine the hypothesis that human reaching movements at different speeds
share similar muscle synergies and to investigate the kinesiology basis and
innervation of muscles. METHODS: Electromyographic activity from six muscles of the upper
limb and shoulder girdle were recorded during three movements at different
speeds, i.e. slow, moderate and fast. The effect of window length on the RMS
signal of the EMG was analyzed and then EMG envelope signals were decomposed
using non-negative matrix factorization. For each of the ten subjects, three
synergies were extracted which accounted for at least 99% of the VAF. For
each movement, the muscle synergies and muscle activation coefficients of
all participants were clustered in to three partitions. Investigation showed
a high similarity and dependency of cluster members due to the cosine
similarity and mutual information in muscle synergy clustering. For further
verification, the EMG envelope signals for all subjects were reconstructed. RESULTS: The results indicated a lower reconstruction error using
the center of the muscle synergy clusters in comparison with the average of
the activation coefficients, which confirms the current research's
hypothesis.
