We model the structural network architecture of the brain as a set of superposed subnetworks, or network components. We use non-negative matrix factorisation, an unsupervised and data-driven approach, to reliably identify separable subnetworks and track their development over the human lifespan. In the NKI-Rockland lifespan sample (n=196), we find evidence for an increased reliance on local communication between neighbouring regions, rather than through heavily-connected network hubs in older age. This method shows good potential for further exploration of the human structural connectome.
This abstract and the presentation materials are available to members only; a login is required.