To evaluate the influence of poststroke aphasia on the functional association of widespread large-scale neuronal networks, we analyzed functional connectivity (FC) between resting-state brain networks (RSNs) in aphasic patients (N = 15) and in healthy volunteers (N = 17) of the same age using resting-state functional connectivity magnetic resonance imaging. As a result, six RSNs were isolated and cross-correlation matrices were computed for their time courses. Aphasic patients showed decreased correlations between posterior part of the default mode (pDMN) and both auditory (AUD) and right frontoparietal (RFP) networks. Additionally, we calculated regions of interest-based FC (ROI-FC), gray and white matter volumes in the ROIs overlapping with pDMN, AUD, and RFP. ROI-FC analysis showed decreased FC between the right pars triangularis and both right middle frontal and right superior frontal gyri. The decreased pDMN-RFP connectivity in patients is likely to reflect changes in FC of these nodes. The lesion in the regions overlapping with pDMN and AUD networks leads to the significantly decreased pDMN-AUD connectivity. Our results suggest that abnormal FC in stroke patients may reflect the impairment of activity not only in the regions directly affected by stroke lesion in the left hemisphere but also in the homotopic regions of the intact right hemisphere. The increase of gray and white matter volume in the right supramarginal gyrus, the functional hub of pDMN, AUD, and RFP networks, correlated with less speech impairment. This increase might reflect a right hemisphere neuroplasticity process to compensate the impaired function of the homotopic region of left frontoparietal network (LFP), pDMN, and AUD in the left hemisphere. The presented results contribute to the hypothesized compensative role of the transfer of attention and executive functions from the damaged areas in the left hemisphere to the right homotopic areas, accompanied by more preserved language skills at the chronic stroke stage.
This work addresses the problem of constructing a unified, topologically optimal connectivity-based brain atlas. The proposed approach aggregates an ensemble partition from individual parcellations without label agreement, providing a balance between sufficiently flexible individual parcellations and intuitive representation of the average topological structure of the connectome. The methods exploit a previously proposed dense connectivity representation, first performing graph-based hierarchical parcellation of individual brains, and subsequently aggregating the individual parcellations into a consensus parcellation. The search for consensus—based on the hard ensemble (HE) algorithm—approximately minimizes the sum of cluster membership distances, effectively estimating a pseudo-Karcher mean of individual parcellations. Computational stability, graph structure preservation, and biological relevance of the simplified representation resulting from the proposed parcellation are assessed on the Human Connectome Project data set. These aspects are assessed using (1) edge weight distribution divergence with respect to the dense connectome representation, (2) interhemispheric symmetry, (3) network characteristics’ stability and agreement with respect to individually and anatomically parcellated networks, and (4) performance of the simplified connectome in a biological sex classification task. Ensemble parcellation was found to be highly stable with respect to subject sampling, outperforming anatomical atlases and other connectome-based parcellations in classification as well as preserving global connectome properties. The HEbased parcellation also showed a degree of symmetry comparable with anatomical atlases and a high degree of spatial contiguity without using explicit priors.