Measuring the Effect of Vision on the Synergy of Lower Extremity Muscles during Walking using Nonnegative Matrix Factorization (NNMF) Algorithm Method

Abstract

Introduction. Lack of visual information in blind people during walking can affect the choice of muscle synergy from among the many incoming messages that reach the central nervous system (CNS). This study aimed to determine the effect of vision on the synergy of lower limb muscles during walking using the nonnegative matrix factorization algorithm (NNMF). Methods. Ten blind people and 10 people with normal vision participated in this study. Activities of involved muscles were recorded during walking. Muscle synergy matrix and synergy activation coefficient were calculated using the NNMF algorithm, while the variance accounted for criterion was used to determine the number of synergies required during walking. In order to assess the similarity of muscle synergy pattern and the relative weight of each muscle in each synergy in each group, Pearson correlation and independent samples t-test at a significance level of α ≤ 0.05 were used. Results. Four muscle synergies were extracted from EMG data during walking. The first (r = 0.431) and the second (r = 0.457) synergy patterns showed a moderate correlation between the two groups. However, the third (r = 0.302) and the fourth (r = 0.329) synergy patterns showed a weak correlation between the two groups. In the blind group, the relative weight of the muscles in the first synergy was significant for the external extensor muscle (P=0.023), and in the second synergy for the biceps femoris. Also, in the third synergy, the relative weight was found to be significant in none of the muscles. In the fourth synergy, however, the relative weight of external extensor muscle in the blind group showed a significant decrease, as compared to the group with normal vision. Conclusions. These changes can be the strategy of the CNS to preserve the optimal functioning in the motor system of blind people.