Interindividual Covariations of Brain Functional and Structural Connectivities Are Decomposed Blindly to Subnetworks: A Fusion-Based Approach

Brain subnetworks most sensitive to alterations of functional connectivity in Schizophrenia: a data-driven approach

Functional connectivity (FC) of the brain changes in various brain disorders. Its complexity, however, makes it difficult to obtain a systematic understanding of these alterations, especially when they are found individually and through hypothesis-based methods. It would be easier if the variety of brain connectivity alterations is extracted through data-driven approaches and expressed as variation modules (subnetworks). In the present study, we modified a blind approach to determine inter-group brain variations at the network level and applied it specifically to schizophrenia (SZ) disorder. The analysis is based on the application of independent component analysis (ICA) over the subject's dimension of the FC matrices, obtained from resting-state functional magnetic resonance imaging (rs-fMRI). The dataset included 27 SZ people and 27 completely matched healthy controls (HC). This hypothesis-free approach led to the finding of three brain subnetworks significantly discriminating SZ from HC. The area associated with these subnetworks mostly covers regions in visual, ventral attention, and somatomotor areas, which are in line with previous studies. Moreover, from the graph perspective, significant differences were observed between SZ and HC for these subnetworks, while there was no significant difference when the same parameters (path length, network strength, global/local efficiency, and clustering coefficient) across the same limited data were calculated for the whole brain network. The increased sensitivity of those subnetworks to SZ-induced alterations of connectivity suggested whether an individual scoring method based on their connectivity values can be applied to classify subjects. A simple scoring classifier was then suggested based on two of these subnetworks and resulted in acceptable sensitivity and specificity with an area under the ROC curve of 77.5%. The third subnetwork was found to be a less specific building block (module) for describing SZ alterations. It projected a wider range of inter-individual variations and, therefore, had a lower chance to be considered as a SZ biomarker. These findings confirmed that investigating brain variations from a modular viewpoint can help to find subnetworks that are more sensitive to SZ-induced alterations. Altogether, our study results illustrated the developed method's ability to systematically find brain alterations caused by SZ disorder from a network perspective.

Combination of structural and functional connectivity explains unique variation in specific domains of cognitive function

The relationship between structural and functional brain networks has been characterised as complex: the two networks mirror each other and show mutual influence but they also diverge in their organisation. This work explored whether a combination of structural and functional connectivity can improve the fit of regression models of cognitive performance. Principal Component Analysis (PCA) was first applied to cognitive data from the Human Connectome Project to identify latent cognitive components: Executive Function, Self-regulation, Language, Encoding and Sequence Processing. A Principal Component Regression approach with embedded Step-Wise Regression (SWR-PCR) was then used to fit regression models of each cognitive domain based on structural (SC), functional (FC) or combined structural-functional (CC) connectivity. Executive Function was best explained by the CC model. Self-regulation was equally well explained by SC and FC. Language was equally well explained by CC and FC models. Encoding and Sequence Processing were best explained by SC. Evaluation of out-of-sample models' skill via cross-validation showed that SC, FC and CC produced generalisable models of Language performance. SC models performed most effectively at predicting Language performance in unseen sample. Executive Function was most effectively predicted by SC models, followed only by CC models. Self-regulation was only effectively predicted by CC models and Sequence Processing was only effectively predicted by FC models. The present study demonstrates that integrating structural and functional connectivity can help explaining cognitive performance, but that the added explanatory value (in sample) may be domain-specific and can come at the expense of reduced generalisation performance (out-of-sample).