Mapping the human brain's cortical-subcortical functional network organization

Abstract

Understanding complex systems such as the human brain requires characterization of the system's architecture across multiple levels of organization-from neurons, to local circuits, to brain regions, and ultimately large-scale brain networks. Here we focus on characterizing the human brain's comprehensive large-scale network organization, as it provides an overall framework for the organization of all other levels. We leveraged the Human Connectome Project dataset to identify network communities across cortical regions, replicating well-known networks and revealing several novel but robust networks, including a left-lateralized language network. We expanded these cortical networks to subcortex, revealing 288 highly-organized subcortical segments that take part in forming whole-brain functional networks. This whole-brain network atlas-released as an open resource for the neuroscience community-places all brain structures across both cortex and subcortex in a single large-scale functional framework, substantially advancing existing atlases to provide a brain-wide functional network characterization in humans.

Task-evoked activity quenches neural correlations and variability across cortical areas

“…In the human fMRI literature, studies have identified large-scale functional brain networks through clustering sets of correlated brain regions using resting-state activity (Ji et al, 2018;Power et al, 2011;Yeo et al, 2011) . During task states, the FC structure has been demonstrated to dynamically reconfigure (Cole et al, 2014;Gonzalez-Castillo and Bandettini, 2017;Krienen et al, 2014) .…”

“…We found that regions that activated more during tasks tend to decrease their global functional FC accordingly during task states. Scatter plots reflect each parcel in the Glasser atlas (Glasser et al, 2016) , and are colored according to network affiliation (Ji et al, 2018) . Best fit lines were estimated using linear regression, but correlations were calculated using a non-parametric rank correlation.…”

The influence of relationship closeness on default-mode network connectivity during social interactions

“…Third, we chose to extract network time-series in this study using a 10-network parcellation as suggested by Smith and colleagues (Smith et al, 2009); we chose this specific parcellation due to its more general anatomical definition of each network and lack of more targeted predictions about subnetworks or network nodes in our task. However, recent work (including studies cited here) has specified more fine-grained parcellations of the DMN and other networks (Dixon et al, 2018;Buckner and DiNicola, 2019;Ji et al, 2019). Future work looking at subnetwork connectivity during social interactions may be better able to characterize specific contributions to encoding of factors such as relationship closeness and outcome value, and may be able to identify how these processes may break down in samples with social difficulties.…”

Neural Correlates of Positive and Negative Valence System Dysfunction in Adolescents Revealed by Data-Driven Parcellation and Resting-State Network Modeling

“…Much of this work has been spearheaded by the Human Connectome Project (HCP) (8), which released a landmark cortical parcellation in 2016 identifying 360 distinct areas based on multimodal measures of cortical thickness, myelination, resting-state functional connectivity, and task activation patterns in an extensively sampled cohort of healthy young adults (9). Recently, the Cole-Anticevic Brain-wide Network Partition (CAB-NP) has extended this parcellation scheme to the subcortex, identifying 358 further regions on the basis of resting-state network assignments and providing a detailed map of discrete functional areas across the entire brain (10). In addition to revealing fundamental aspects of neural organization, high-quality parcellations provide an invaluable framework for further data-driven research.…”

“…As clinical symptomatology is salient across disorders and lies on a continuum within each disorder, our study was designed to capture the full range of symptom severity by recruiting a large, transdiagnostic sample that included adolescents with comorbid and subthreshold diagnoses as well as healthy controls. Using graph theory, we examined relationships between clinical symptomatology and resting-state network properties of centrality and efficiency within the functionally accurate CAB-NP network (10). As reward circuitry plays a central role in the emergence of psychiatric conditions during adolescence, we also repeated analyses within three functionally defined reward networks derived from the Reward Flanker Task (RFT) (15).…”