Dynamic reconfiguration of frontal brain networks during executive cognition in humans

The brain is an inherently dynamic system, and executive cognition requires dynamically reconfiguring, highly evolving networks of brain regions that interact in complex and transient communication patterns. However, a precise characterization of these reconfiguration processes during cognitive function in humans remains elusive. Here, we use a series of techniques developed in the field of “dynamic network neuroscience” to investigate the dynamics of functional brain networks in 344 healthy subjects during a working-memory challenge (the “n-back” task). In contrast to a control condition, in which dynamic changes in cortical networks were spread evenly across systems, the effortful working-memory condition was characterized by a reconfiguration of frontoparietal and frontotemporal networks. This reconfiguration, which characterizes “network flexibility,” employs transient and heterogeneous connectivity between frontal systems, which we refer to as “integration.” Frontal integration predicted neuropsychological measures requiring working memory and executive cognition, suggesting that dynamic network reconfiguration between frontal systems supports those functions. Our results characterize dynamic reconfiguration of large-scale distributed neural circuits during executive cognition in humans and have implications for understanding impaired cognitive function in disorders affecting connectivity, such as schizophrenia or dementia.The era of human brain mapping has demonstrated the power of associating brain regions to specific cognitive functions. However, emerging evidence indicates that many so-called “domain-general” areas engage in multiple functions, differing from “domain-specific” areas such as primary visual cortex that perform a very specific function (1, 2). Such broad engagement is enabled by two fundamental features of brain function: time and interconnectivity. Brain areas and associated circuits or networks may be engaged in tasks differently over time: some transiently and some consistently (2, 3). A fundamental understanding of cognition in general and executive cognition in particular should therefore address the dynamic, interconnected nature of brain function.Here, we use and extend emerging tools from “dynamic network neuroscience,” a field of neuroscientific inquiry that embraces the inherently evolving, interconnected nature of neurophysiological phenomena underlying human cognition (3, 4). Building on the formalism of network science (5), this approach treats the patterns of communication between brain regions as evolving networks and links this evolution to behavioral outcomes. Conceptually, this approach is particularly useful in examining the consistent or transient engagement of neural (or cognitive) circuits or putative functional modules (Fig. 1). We define a network module to be a set of brain regions that are strongly connected to each other and weakly connected to the rest of the network. Using dynamic network-based clustering techniques (6), we seek to observe the flexible recruitment and integration of neural circuits underlying executive function in the form of working memory.Open in a separate windowFig. 1.Network reconfiguration during executive function. (A) We use a numerical n-back task consisting of 0-back and 2-back conditions. (B) We define 270 cortical and subcortical regions of interest (36), and (C) extract the mean time course from each region. (D) A sliding window comprising 15 volumes with no gap was applied to regional mean time courses, and for each window we estimated the functional connectivity between pairs of regions using coherence. This procedure resulted in a sequence of 114 time-ordered adjacency matrices. (E) Using a dynamic community detection algorithm (part 1 in panel), we identified network modules in each time window and tracked their evolution over time. (F) By estimating the probability that a brain region changes its allegiance to modules between any two consecutive time windows (part 2 in panel), we observed that whole-brain flexibility oscillated between unitask (2-back or 0-back only) and dual-task (2-back and 0-back in same time window) conditions.Working memory lies at the interface of perception and action (7) and requires the integration of large-scale neural circuits (8–11). Theoretical frameworks for working memory call on the interplay of distinct components (12) and their integration in broader cognitive circuits (1). The empirical neuroimaging literature has bolstered these conceptualizations by identifying several distinct sets of brain areas underlying working-memory performance (13–17). Nevertheless, a fundamental understanding of the flexible integration and recruitment of these circuits remains incomplete.In the present study, we characterize the time-dependent interactions between putative neural circuits [network modules (3)] underlying working-memory performance in humans as elicited by an n-back task performed during the acquisition of functional MRI (fMRI) data (Fig. 1 A and B). By deploying a sliding time window analysis (18, 19), we capture brain network dynamics during working-memory function (2-back), during a baseline condition (0-back), and in transitions between baseline and task (Fig. 1 C and D). We identify putative functional modules in each time window and track how brain regions change their engagement in these modules over time (Fig. 1 E and F). We quantify those changes over time by flexibility, which measures how often a particular brain region changes its modular allegiance. Based on the cognitive load of the 2-back condition (20–22), we hypothesize that the brain transiently reorganizes functional modules during task performance in comparison with baseline. Furthermore, we hypothesize that this reconfiguration is driven by higher order cognitive control systems, particularly in frontal cortex (2), which are known to play a role in task switching. Finally, based on prior evidence linking network reconfiguration to behavioral adaptation (3), we hypothesize that individuals who display more flexible network structures will perform better than individuals with more rigid network structures.