Muscle synergies control during hand-reaching tasks in multiple directions post-stroke

Sharon Israely, Gerry Leisman, Chay C. Machluf, Eli Carmeli

Research output: Contribution to journalArticlepeer-review


Purpose: A muscle synergies model was suggested to represent a simplifying motor control mechanism by the brainstem and spinal cord. The aim of the study was to investigate the feasibility of such control mechanisms in the rehabilitation of post-stroke individuals during the execution of hand-reaching movements in multiple directions, compared to non-stroke individuals. Methods: Twelve non-stroke and 13 post-stroke individuals participated in the study. Muscle synergies were extracted fromEMGdata that was recorded during hand reaching tasks, using the NMF algorithm. The optimal number of synergies was evaluated in both groups using the Variance Accounted For (VAF) and the Mean Squared Error (MSE). A cross validation procedure was carried out to define a representative set of synergies. The similarity index and the K-means algorithm were applied to validate the existence of such a set of synergies, but also to compare the modulation properties of synergies for different movement directions between groups. The similarity index and hierarchical cluster analysis were also applied to compare between group synergies. Results: Four synergies were chosen to optimally capture the variances in the EMG data, with mean VAF of 0.917 ± 0.034 and 0.883 ± 0.046 of the data variances, with respective MSE of 0.007 and 0.016, in the control and study groups, respectively. The representative set of synergies was set to be extracted from movement to the center of the reaching space. Two synergies had different muscle activation balance between groups. Seven and 17 clusters partitioned the muscle synergies of the control and study groups. The control group exhibited a gradual change in the activation in the amplitude in the time domain (modulation) of synergies, as reflected by the similarity index, whereas the study group exhibited consistently significant differences between all movement directions and the representative set of synergies. The study findings support the existence of a representative set of synergies, which are modulated to execute movements in different directions. Conclusions: Post-stroke individuals differently modulate the activation of synergies to different movement directions than do non-stroke individuals. The conclusion was supported by different muscle activation balances, similarity values and different classifications of synergies among groups.

Original languageEnglish
Article number10
JournalFrontiers in Computational Neuroscience
StatePublished - 23 Feb 2018

Bibliographical note

Funding Information:
The authors thank Nimrod Peleg, Tal Shnitzer, Prof. Ronen Talmon and Jonathan Brokman of the Faculty of Electrical Engineering, at the Technion Israel Institute of Technology, Haifa, Israel for their professional consultation regarding the development of the algorithm.

Publisher Copyright:
© 2018 Israely, Leisman, Machluf and Carmeli.


  • Electromyography (EMG)
  • Hand reaching
  • Modulation
  • Motor control
  • Muscle synergies
  • Non-negative matrix factorization (NMF)
  • Post stroke
  • Variance accounted for (VAF)

ASJC Scopus subject areas

  • Cellular and Molecular Neuroscience
  • Neuroscience (miscellaneous)


Dive into the research topics of 'Muscle synergies control during hand-reaching tasks in multiple directions post-stroke'. Together they form a unique fingerprint.

Cite this