Towards understanding interindividual differences in cortical morphological brain networks

Individual-level morphological brain networks are becoming an important approach for studying human connectome; however, their interindividual differences are not well understood with respect to behavioral and cognitive relevance, individual identification, and genetic origin. Using three publicly available datasets that involved cross-sectional and longitudinal structural magnetic resonance scans of adults and children, we constructed four morphological brain networks for each of 1,451 images from 1,329 participants on the basis of cerebral surface-based, vertex-wise cortical thickness, fractal dimension, gyrification index and sulcal depth, respectively. The morphological index-dependent networks were further fused via multiplex network model, and fed into community detection. We found that the multiplex morphological brain networks 1) accounted for significant proportions of interindividual variance in and were predictive of multiple behavioral and cognitive domains, in particular Cognition and Motor domains (P < 0.05, corrected), 2) distinguished individuals from each other with high accuracies even for twin subjects (accuracies > 96%), and 3) exhibited low-moderate heritability with the highest for sulcal depth-based morphological brain networks. Intriguingly, compared with intra-module morphological connectivity, inter-module connections explained more behavioral and cognitive variance and were associated with higher heritability. Further comparisons revealed that multiplex morphological brain networks outperformed each type of single-layer morphological brain networks in the performance of behavioral and cognitive association and prediction, and individual identification. Finally, all the findings were generally reproducible over different datasets. Altogether, our findings indicate that interindividual differences in individual-level morphological brain networks are biologically meaningful, which underpins their usage as fingerprints for individualized studies in health and disease.