Muscle synergy analysis of lower limb based on Mechanomyography

Abstract

Analysis of muscle synergy during the execution of different motions can provide a physiological basis for the rehabilitation assessment of stroke patients. Mechanomyography (MMG) signal is a kind of low-frequency signal produced during muscle vibration, has been widely applied to pattern recognition and muscle fatigue estimation. In this paper, muscle synergy was extracted from 5 channels of MMG signals recorded from 8 healthy subjects in the lower limbs using the non-negative matrix factorization (NNMF) algorithm. In addition, the similarities of muscle activation patterns of 4 different motions were analyzed, and a suitable activation threshold was selected by comparing synergistic and non-synergistic muscles through the coherence analysis method. This study provides a scientific basis for studying muscle synergy based on MMG signals.