Correlation Analysis of Surface Electromyography Signals Based on Mapping Graph Structure

In rehabilitation medicine, surface electromyography (sEMG) is a non-invasive measurement of muscle activity that is widely researched. Current research on sEMG often uses deep learning and other techniques to analyze its Temporal regularity, but ignores the relationship between multiple sampling points. In this paper, a correlation analysis algorithm of sEMG signals between multiple sampling points is proposed. By analyzing the spatial relationship between sampling points, a spatiotemporal structure diagram is constructed, and a correlation analysis matrix is established to visualize the correlation degree of different sampling points. The correctness of the correlation analysis is verified in the public dataset. In future research, the graph neural network will be used to capture the degree of correlation between sampling points, so as to analyze the sEMG signals of multiple sampling points more comprehensively and accurately.