Fig. 9: Inter-muscular coherence networks for neonates and adults.

a First two components (accounting on average 79% of VAF, 95% confidence interval [69%, 86%]) obtained with non-negative matrix factorization of the frequency content of the coherence spectra of all muscle combinations in neonates (n = 13). The components are ordered (from top to bottom) based on the peak of the frequency components (c1–c2, left column, black lines denote average across subjects +SD in dashed black lines). The central column shows the average loadings (w1–w2) of the corresponding frequency component across subjects. These loadings give the strength of the edges between the eight nodes of each muscle-weighted network (n1–n2, right column), connection strength is reflected by the width of the lines. b First four components (accounting on average 82% of VAF, 95% confidence interval [78%, 86%]) obtained with non-negative matrix factorization of the frequency content of the coherence spectra of all muscle combinations in adults (n = 15, same format as a). c Summary statistics of inter-muscular coherence networks. Network metrics were used to statistically compare the muscle networks across groups and frequencies: average (±SD) 50% frequency band across subjects (upper row, left panel), average (+SD) betweenness-centrality across subjects (upper row, right panel), and average (+SD) edge weight across subjects (lower row) considering all network edges (global, left panel), only the intra-limb edges (intra-limb, central panel) or only the interlimb edges (interlimb, right panel). Data points for all individuals are included in the histograms. Horizontal black lines denote significant differences between networks (post-hoc Tukey-Kramer multiple comparison p < 0.05).