Adaptive Reconfiguration of Intrinsic Community Structure in Children with 5-Year Abacus Training.
Yi Zhang,Chunjie Wang,Yuzhao Yao,Changsong Zhou,Feiyan Chen
DOI: https://doi.org/10.1093/cercor/bhab010
IF: 4.861
2021-01-01
Cerebral Cortex
Abstract:Human learning can be understood as a network phenomenon, underpinned by the adaptive reconfiguration of modular organization. However, the plasticity of community structure (CS) in resting-state network induced by cognitive intervention has never been investigated. Here, we explored the individual difference of intrinsic CS between children with 5-year abacus-based mental calculation (AMC) training (35 subjects) and their peers without prior experience in AMC (31 subjects). Using permutation-based analysis between subjects in the two groups, we found the significant alteration of intrinsic CS, with training-attenuated individual difference. The alteration of CS focused on selective subsets of cortical regions ("core areas"), predominantly affiliated to the visual, somatomotor, and default-mode subsystems. These subsystems exhibited training-promoted cohesion with attenuated interaction between them, from the perspective of individuals' CS. Moreover, the cohesion of visual network could predict training-improved math ability in the AMC group, but not in the control group. Finally, the whole network displayed enhanced segregation in the AMC group, including higher modularity index, more provincial hubs, lower participation coefficient, and fewer between-module links, largely due to the segregation of "core areas." Collectively, our findings suggested that the intrinsic CS could get reconfigured toward more localized processing and segregated architecture after long-term cognitive training.