Atlasing white matter and grey matter joint contributions to resting-state networks in the human brain

Over the past two decades, the study of resting-state functional magnetic resonance imaging has revealed that functional connectivity within and between networks is linked to cognitive states and pathologies. However, the white matter connections supporting this connectivity remain only partially described. We developed a method to jointly map the white and grey matter contributing to each resting-state network (RSN). Using the Human Connectome Project, we generated an atlas of 30 RSNs. The method also highlighted the overlap between networks, which revealed that most of the brain's white matter (89%) is shared between multiple RSNs, with 16% shared by at least 7 RSNs. These overlaps, especially the existence of regions shared by numerous networks, suggest that white matter lesions in these areas might strongly impact the communication within networks. We provide an atlas and an open-source software to explore the joint contribution of white and grey matter to RSNs and facilitate the study of the impact of white matter damage to these networks. In a first application of the software with clinical data, we were able to link stroke patients and impacted RSNs, showing that their symptoms aligned well with the estimated functions of the networks. A grey and white atlas of resting-state networks and open-source software to access are presented and applied to identify impacted networks in stroke patients.

Automated summary

Resting-state functional magnetic resonance imaging has shown that functional connectivity within and between networks is related to cognitive states and pathologies. However, the white matter connections supporting this connectivity have only been partially described. A new method has been developed to map the white and grey matter contributing to each resting-state network, demonstrating the overlap between networks and the potential impact of white matter lesions on network communication. An atlas and open-source software have been provided to facilitate the study of white matter damage to resting-state networks, with initial application showing promising results in identifying impacted networks in stroke patients.