Fig 1.
Neurocognitive intrinsic connectivity networks.
The spatial maps of neurocognitive networks (thresholded at Z > 3) shown here represent the anterior, posterior and ventral subnetworks of the default mode network (a/p/v DMN), and the left and right frontoparietal (l/r FPN), dorsal attentional (DAN), cingulo-opercular (CON), and frontopolar (FN) networks. The spatial maps are displayed in sagittal and axial views and superimposed on the MNI152 standard space template image.
Fig 2.
Group differences in within-network FC of the ICNs during resting state and tasks.
Using combined group ICNs as templates for dual regression, the between groups comparison showed that (a) during resting state, children compared to adults, had a significantly stronger FC in the ACC within the CON1. (b) During tasks, children compared to adults had stronger FC in five ICNs representing the aDMN, vDMN, rFPN, CON1 and CON2. All comparison analyses were FDR corrected at p < 0.05 with a cluster size of > 10 contiguous voxels. The statistical maps of significant FC differences are presented in green, and the corresponding group ICN templates are shown in red/yellow. All statistical maps are displayed on selected slice planes of the MNI152 standard brain template. The columns illustrate the mean FC within the area that differed significantly between the groups. ACC, anterior cingulate cortex; CG, cingulate gyrus; CON, cingulo-opercular network; DMN, default mode network; FPN, frontoparietal network; IFG, inferior frontal gyrus; MFG, middle frontal gyrus; PFC, prefrontal cortex; RC, retrosplenial cortex; SFG, superior frontal gyrus; a, anterior; v, ventral; med, medial part; R, r, right; *** p < 0.001 (unpaired t-test).
Fig 3.
Hierarchical clustering, full and partial correlations and group differences in between-network FC of the ICNs.
Hierarchical clustering of the neurocognitive networks of the combined group data during (a) resting state and (b) task performance. Full and partial correlations between the ICNs representing neurocognitive networks are displayed below and above the main diagonal, respectively. (c) Adults compared to children showed stronger connectivity between the aDMN and rFPN during tasks (indicated by letter c in panel b). (d) Children compared to adults had stronger connectivity between the aDMN and the frontopolar network during tasks (indicated by letter d in panel b). The significance level of all comparison analyses was multiple comparison corrected at p < 0.05 with cluster size > 10 contiguous voxels. The columns in c and d illustrate the mean FC between the networks that differed significantly between the groups. CON, cingulo-opercular network; DAN, dorsal attentional network; DMN, default mode network; FN, frontopolar network; FPN, frontoparietal network; a, anterior; p, posterior; v, ventral; l, left; R, r, right; ** p < 0.01, *** p < 0.001 (unpaired t-test).
Fig 4.
Group difference in FC strength of the rFPN between resting state and tasks.
The rFPN showed significant group x brain state interaction of FC in the repeated measures ANOVA. The mean FC within the rFPN was significantly stronger in children than adults during tasks. The columns illustrate the average z-score across voxels in the rFPN. FPN, frontoparietal network; RS, resting state; T, tasks; R, r, right; ** p < 0.01 (two-way repeated measures ANOVA, followed by unpaired t-tests).
Table 1.
The ICNs that showed stronger FC in children than adults in the task fMRI data.
Fig 5.
Correlations between age, behavior and FC during tasks.
In children, accuracy of the task performance correlated positively with age (a), and the FC strength within the aDMN (the mean z-value of the aDMN network) correlated negatively with age (b) and positively with RT (c). In adults (d), the task performance accuracy correlated negatively with the between-network connectivity. CON, cingulo-opercular network; DMN, default mode network; a, anterior; v, ventral.
Table 2.
The response times and accuracy (d’) in children and adults during tasks.