Fig. 1: The cortical variation of between-hemisphere functional distance.

a To compute between-hemisphere functional divergence, we first calculated functional connectivity (FC) between vertices within each hemisphere based on vertexwise time series data. Then we applied the diffusion map embedding method on hemispheric FC matrices to obtain the functional gradients. These gradients form a low-dimensional representation of the original functional connectivity space, in which the distance represents the functional similarity between vertices. To ensure better comparability between homotopic vertices within the representational spaces of the left and right hemispheres, we employed functional alignment which brought the individual left and right hemisphere embeddings into a group-level common functional space. The final between-hemisphere functional divergence was measured as the Euclidean distance between each pair of homotopic vertices in a 6-dimensional common representational space. b The group average map of between-hemisphere functional distance in the HCP datasets (N = 777). c To further investigate between-hemisphere functional divergence at the network level, we plotted the functional distances based on the Yeo 7-network parcellation. A set of vertices that belonged to distinct networks across the two hemispheres was defined as the mismatch zone. The whiskers in the boxplot represent data within 1.5 times the interquartile range.