Network Control Theory in Cognitive Neuroscience

Network Control Theory applies mathematical control principles to structural brain networks, offering mechanistic insight into how brain regions steer neural dynamics between cognitive states. By representing the connectome as a graph of nodes linked by white-matter tracts, this framework quantifies the ease or difficulty of transitioning from one pattern of activation to another. Central to this approach is the concept of controllability, which captures the capacity of individual regions to influence whole-brain dynamics, and the notion of energy, denoting the input required to induce a desired state transition. Emerging evidence has shown that densely connected hub regions facilitate effortless shifts to common cognitive states, while control hubs in frontoparietal systems enable navigation to more challenging or novel states. This paradigm has illuminated how developmental and age-related changes in network topology may support or hinder cognitive flexibility, and has provided a quantitative basis for understanding individual differences in mental health and treatment response. Moreover, integration of linear and nonlinear models has revealed that, at macroscopic scales, relatively simple linear approximations may suffice to describe resting-state dynamics, greatly simplifying the mathematical treatment. Practical applications span optimisation of brain stimulation targets, prediction of cognitive decline, design of personalised interventions in psychiatric disorders, and the interpretation of pharmacological modulation. By uniting graph theory, control theory and neuroimaging, Network Control Theory has matured into a versatile framework for probing the architecture and dynamics of the human brain.

Research from Nature Portfolio

In the past two years, researchers have harnessed advanced diffusion and functional imaging to refine models of brain network control. One study has demonstrated that linear autoregressive frameworks can accurately capture resting-state dynamics across hundreds of subjects, suggesting that macroscopic neural activity may be governed by processes amenable to linear control strategies. Another investigation into healthy ageing has revealed that average controllability declines within key networks such as frontoparietal control and default mode, but that topological redundancy can compensate for these deficits, preserving cognitive function in later life. Foundational work has also delineated distinct control roles for densely connected default mode hubs, boundary regions mediating integration, and weakly connected control nodes, establishing the mechanistic basis for transitions among diverse cognitive states.

Network Control Theory in Cognitive Neuroscience publication trend

The graph below shows the total number of articles in network control theory in cognitive neuroscience across all publications each year (not limited to Nature Index journals).

Technical terms

Network controllability: A measure of the capacity of a brain region to steer the system into different states through external or internal inputs.

Average controllability: The ease with which low-energy inputs at a given node can move the network into a variety of nearby states.

Structural connectivity: The anatomical map of white-matter pathways linking brain regions, typically inferred from diffusion MRI.

Functional connectivity: Statistical relationships between neural time series that reflect synchrony or coordination among brain regions.

Network topology: The organisation of nodes and edges in a network, determining patterns of integration, segregation and redundancy.

References

1. Controllability of structural brain networks. Nature Communications (2015).
2. Developmental increases in white matter network controllability support a growing diversity of brain dynamics. Nature Communications (2017).
3. Stimulation-Based Control of Dynamic Brain Networks. PLOS Computational Biology (2016).
4. Genetic, individual, and familial risk correlates of brain network controllability in major depressive disorder. Molecular Psychiatry (2023).
5. Macroscopic resting-state brain dynamics are best described by linear models. Nature Biomedical Engineering (2023).
6. Age-related differences in network controllability are mitigated by redundancy in large-scale brain networks. Communications Biology (2024).