Revealing the relevant spatiotemporal scale underlying whole-brain dynamics

Abstract

The brain can rapidly process and transfer information between cortical brain networks by dynamically transitioning between brain states. Here we study the switching activity between functional brain networks that have been estimated at various spatial scales from n = 100 to n = 1000 using resting-state fMRI data. We also generate timeseries at different temporal scales from milliseconds to seconds using whole-brain modelling. We calculate the entropy of switching activity between functional brain networks which represents the richness of the dynamical repertoire. We provide evidence that the entropy of functional network occurrence follows an inverted U-shaped curve with a maximum at a spatial scale between 200 and 300 regions and at a temporal scale of 200 milliseconds.

Neural mass modelling for the masses: Democratising access to whole-brain biophysical modelling with FastDMF

“…The existent implementations of the DMF are based on parcellations with less than 100 regions, –typically given by the Automatic Anatomical Labelling (AAL90) 36 or the Desikan-Killiany 37 parcellation– because, otherwise, it would be unfeasible to simulate brain activity based on fine-grained parcellations. However, recent advances in neuroimaging data analysis involve measuring brain activity at multiple spatial scales, from coarse to fine-grained, revealing new insights about underlying brain dynamics 38 and its relevant operational scales 39 . Therefore, being restricted to simulating biophysically realistic brain activity at a coarse-grained spatial scale represents another barrier that hinders the ability of neuroscientists to leverage the full potential of DMF modelling.…”