Brain parcellation driven by dynamic functional connectivity better capture intrinsic network dynamics

Abstract

Until now, dynamic functional connectivity (dFC) based on functional magnetic resonance imaging is typically estimated on a set of predefined regions of interest (ROIs) derived from an anatomical or static functional atlas which follows an implicit assumption of functional homogeneity within ROIs underlying temporal fluctuation of functional coupling, potentially leading to biases or underestimation of brain network dynamics. Here, we presented a novel computational method based on dynamic functional connectivity degree (dFCD) to derive meaningful brain parcellations that can capture functional homogeneous regions in temporal variance of functional connectivity. Several spatially distributed but functionally meaningful areas that are well consistent with known intrinsic connectivity networks were identified through independent component analysis (ICA) of time‐varying dFCD maps. Furthermore, a systematical comparison with commonly used brain atlases, including the Anatomical Automatic Labeling template, static ICA‐driven parcellation and random parcellation, demonstrated that the ROI‐definition strategy based on the proposed dFC‐driven parcellation could better capture the interindividual variability in dFC and predict observed individual cognitive performance (e.g., fluid intelligence, cognitive flexibility, and sustained attention) based on chronnectome. Together, our findings shed new light on the functional organization of resting brains at the timescale of seconds and emphasized the significance of a dFC‐driven and voxel‐wise functional homogeneous parcellation for network dynamics analyses in neuroscience.

A detailed analysis of anatomical plausibility of crossed and uncrossed streamline rendition of the dentato-rubro-thalamic tract (DRT(T)) in a commercial stereotactic planning system

“…As a consequence, the tractographic rendition of the DRT represents only a part of this network [ 13 – 15 ]. There is emerging evidence that the DRT represents the very part of the tremor network which is accessible to therapeutic approaches like deep brain stimulation (DBS) and stereotactic lesion surgery (SLS), including MR-guided focused ultrasound (MRgFUS) [ 1 , 3 , 8 , 12 , 14 , 15 , 17 , 21 , 29 , 32 , 33 , 39 , 60 , 64 , 70 ]. Still, some of this work is retrospective in nature [ 1 , 14 , 39 , 51 , 55 ] and shows thalamic connectivity related to tremor improvement rather than the DRT as structure itself.…”

“…Naturally, such synapses along with the (activation) direction cannot be displayed with the dMRI technology. A surgically and streamline-anatomically more important feature is the definition of the thalamic penetration level at the Vim nucleus with tractographic methods [ 1 , 14 , 57 , 64 ] since the Vim nucleus and the adjacent subthalamic level are the classical target region [ 46 ] for any tremor reducing surgery [ 17 , 21 ] especially in essential tremor (ET) and Parkinson’s disease (PD) tremor. These structures cannot directly be visualized in conventional anatomical sequences.…”

Revealing the Relevant Spatiotemporal Scale Underlying Whole-Brain Dynamics

“…Although these results are promising, they focused on separate brain regions only. In contrast, we were explicitly interested in whole-brain networks, as functional networks seem to be closer related to cognitive processes than single brain regions (Dadi et al, 2019;Fan et al, 2021).…”

Unsupervised clustering of track-weighted dynamic functional connectivity reveals white matter substrates of functional connectivity dynamics

“…ICA decomposition was performed at three different dimensionality levels, by selecting a different number of components (n) for each run: a first run with n=10 (ICA10), to reveal large-scale networks; a second run with n=20 (ICA20) which has been commonly used empirically to identify consistent resting-state networks 7,26 and a third run with n=100 (ICA100) to obtain a more fine-grained parcellation. For each of these three group ICA runs, and both for the principal dataset (HCP) and the validation dataset (LEMON), peak activation for each independent component (IC) was localized in the white matter; all the components' mean power spectra show higher low-frequency spectral power and for all the reconstructed components there was no or minimal overlap with known vascular, ventricular, and meningeal sources of artifacts (Supplementary File 1).…”

“…Recently, this paradigm has been further expanded by incorporating evidence for time-varying fluctuations in functional connectivity strength across brain regions 6 . This dynamic functional connectivity approach has been employed to parcellate brain into brain networks akin to those identified by static functional connectivity 7 , as well as to identify transitions from different states of brain activity in the resting state 8 , and to predict inter-individual variability in age and cognition 9,10 .…”