Similarity and stability of face network across populations and throughout adolescence and adulthood

Exploring continual learning strategies in artificial neural networks through graph-based analysis of connectivity: Insights from a brain-inspired perspective

Artificial Neural Networks (ANNs) aim at mimicking information processing in biological networks. In cognitive neuroscience, graph modeling is a powerful framework widely used to study brain structural and functional connectivity. Yet, the extension of graph modeling to ANNs has been poorly explored especially in terms of functional connectivity (i.e. the contextual change of the activity's units in networks). In the perspective of designing more robust and interpretable ANNs, we study how a brain-inspired graph-based approach can be extended and used to investigate ANN properties and behaviors. We focus our study on different continual learning strategies inspired by the biological mechanisms and modeled with ANNs. We show that graph modeling offers a simple and elegant framework to deeply investigate ANNs, compare their performances, and explore deleterious behaviors such as catastrophic forgetting.

How the intrinsic functional connectivity patterns of the semantic network support semantic processing

Semantic processing, a core of language comprehension, involves the activation of brain regions dispersed extensively across the frontal, temporal, and parietal cortices that compose the semantic network. To comprehend the functional structure of this semantic network and how it prepares for semantic processing, we investigated its intrinsic functional connectivity (FC) and the relation between this pattern and semantic processing ability in a large sample from the Human Connectome Project (HCP) dataset. We first defined a well-studied brain network for semantic processing, and then we characterized the within-network connectivity (WNC) and the between-network connectivity (BNC) within this network using a voxel-based global brain connectivity (GBC) method based on resting-state functional magnetic resonance imaging (fMRI). The results showed that 97.73% of the voxels in the semantic network displayed considerably greater WNC than BNC, demonstrating that the semantic network is a fairly encapsulated network. Moreover, multiple connector hubs in the semantic network were identified after applying the criterion of WNC > 1 SD above the mean WNC of the semantic network. More importantly, three of these connector hubs (i.e., the left anterior temporal lobe, angular gyrus, and orbital part of the inferior frontal gyrus) were reliably associated with semantic processing ability. Our findings suggest that the three identified regions use WNC as the central mechanism for supporting semantic processing and that task-independent spontaneous connectivity in the semantic network is essential for semantic processing.

Hemispheric asymmetry in cortical thinning reflects intrinsic organization of the neurotransmitter systems and homotopic functional connectivity

Hemispheric lateralization and its origins have been of great interest in neuroscience for over a century. The left-right asymmetry in cortical thickness may stem from differential maturation of the cerebral cortex in the two hemispheres. Here, we investigated the spatial pattern of hemispheric differences in cortical thinning during adolescence, and its relationship with the density of neurotransmitter receptors and homotopic functional connectivity. Using longitudinal data from IMAGEN study (N = 532), we found that many cortical regions in the frontal and temporal lobes thinned more in the right hemisphere than in the left. Conversely, several regions in the occipital and parietal lobes thinned less in the right (vs. left) hemisphere. We then revealed that regions thinning more in the right (vs. left) hemispheres had higher density of neurotransmitter receptors and transporters in the right (vs. left) side. Moreover, the hemispheric differences in cortical thinning were predicted by homotopic functional connectivity. Specifically, regions with stronger homotopic functional connectivity showed a more symmetrical rate of cortical thinning between the left and right hemispheres, compared with regions with weaker homotopic functional connectivity. Based on these findings, we suggest that the typical patterns of hemispheric differences in cortical thinning may reflect the intrinsic organization of the neurotransmitter systems and related patterns of homotopic functional connectivity.

Functional connectomes become more longitudinally self-stable, but not more distinct from others, across early childhood

Functional connectomes, as measured with functional magnetic resonance imaging (fMRI), are highly individualized, and evidence suggests this individualization may increase across childhood. A connectome can become more individualized either by increasing self-stability or decreasing between-subject-similarity. Here we used a longitudinal early childhood dataset to investigate age associations with connectome self-stability, between-subject-similarity, and developmental individualization, defined as an individual's self-stability across a 12-month interval relative to their between-subject-similarity. fMRI data were collected during an 18-minute passive viewing scan from 73 typically developing children aged 4-7 years, at baseline and 12-month follow-up. We found that young children had highly individualized connectomes, with sufficient self-stability across 12-months for 98% identification accuracy. Linear models showed a significant relationship between age and developmental individualization across the whole brain and in most networks. This association appeared to be largely driven by an increase in self-stability with age, with only weak evidence for relationships between age and similarity across participants. Together our findings suggest that children's connectomes become more individualized across early childhood, and that this effect is driven by increasing self-stability rather than decreasing between-subject-similarity.