Structural-Functional Connectivity Bandwidth of the Human Brain
Parsons, Nicholas, Ugon, Julien, Morgan, Kerri, Shelyag, Sergiy, Hocking, Alex, Chan, Su, Poudel, Govinda, Dominguez Duque, Juan and Caeyenberghs, Karen. (2022). Structural-Functional Connectivity Bandwidth of the Human Brain. NeuroImage. 263, pp. 1-12. https://doi.org/10.1016/j.neuroimage.2022.119659
|Authors||Parsons, Nicholas, Ugon, Julien, Morgan, Kerri, Shelyag, Sergiy, Hocking, Alex, Chan, Su, Poudel, Govinda, Dominguez Duque, Juan and Caeyenberghs, Karen|
Background: The human brain is a complex network that seamlessly manifests behaviour and cognition. This network comprises neurons that directly, or indirectly mediate communication between brain regions. Here, we show how multilayer/multiplex network analysis provides a suitable framework to uncover the throughput of structural connectivity (SC) to mediate information transfer-giving rise to functional connectivity (FC).
Method: We implemented a novel method to reconcile SC and FC using diffusion and resting-state functional MRI connectivity data from 484 subjects (272 females, 212 males; age = 29.15 ± 3.47) from the Human Connectome Project. First, we counted the number of direct and indirect structural paths that mediate FC. FC nodes with indirect SC paths were then weighted according to their least restrictive SC path. We refer to this as SC-FC Bandwidth. We then mapped paths with the highest SC-FC Bandwidth across 7 canonical resting-state networks.
Findings: We found that most pairs of FC nodes were connected by SC paths of length two and three (SC paths of length >5 were virtually non-existent). Direct SC-FC connections accounted for only 10% of all SC-FC connections. The majority of FC nodes without a direct SC path were mediated by a proportion of two (44%) or three SC path lengths (39%). Only a small proportion of FC nodes were mediated by SC path lengths of four (5%). We found high-bandwidth direct SC-FC connections show dense intra- and sparse inter-network connectivity, with a bilateral, anteroposterior distribution. High bandwidth SC-FC triangles have a right superomedial distribution within the somatomotor network. High-bandwidth SC-FC quads have a superoposterior distribution within the default mode network.
Conclusion: Our method allows the measurement of indirect SC-FC using undirected, weighted graphs derived from multimodal MRI data in order to map the location and throughput of SC to mediate FC. An extension of this work may be to explore how SC-FC Bandwidth changes over time, relates to cognition/behavior, and if this measure reflects a marker of neurological injury or psychiatric disorders.
|Year||01 Jan 2022|
|Journal citation||263, pp. 1-12|
|Digital Object Identifier (DOI)||https://doi.org/10.1016/j.neuroimage.2022.119659|
|Web address (URL)||https://www.sciencedirect.com/science/article/pii/S1053811922007807|
|Open access||Published as green open access|
|Research or scholarly||Research|
File Access Level
|08 Oct 2022|
|Publication process dates|
|Accepted||29 Sep 2022|
|Deposited||10 Jan 2023|
© 2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
|Place of publication||Netherlands|
|License: CC BY-NC-ND 4.0|
|File access level: Open|
4views this month
3downloads this month