Controlling target brain regions by optimal selection of input nodes.

Area of Science:

• Neuroscience
• Network Science
• Computational Biology

Background:

• Network control theory offers potential for brain stimulation but faces limitations in modeling brain dynamics and controlling whole-brain activity.
• Existing frameworks often use simplified models, hindering precise application in neurostimulation experiments.

Purpose of the Study:

• To refine network control theory for practical neurostimulation by focusing on the control of specific brain regions or subnetworks.
• To investigate the feasibility and optimal strategies for low-energy, targeted brain control using individual-level effective connectivity.

Main Methods:

• Leveraged recent advancements in linear modeling of brain dynamics, specifically effective connectivity.
• Applied the concept of target controllability to analyze the control of single regions or small subnetworks.
• Utilized the asymmetric effective connectome, contrasting with previous research using the symmetric structural connectome.

Main Results:

• Identified conditions for achieving targeted brain control with low energy cost and minimal stimulation points.
• Provided general predictions for optimal stimulation locations based on the desired subset of controlled brain regions.
• Highlighted the distinct roles of in- and out-hubs and the importance of inhibitory connections in effective brain control, using the asymmetric connectome.

Conclusions:

• Effective brain control strategies should be individualized due to significant inter-individual variation in effective connectomes.
• The refined approach using effective connectivity offers a more precise framework for designing neurostimulation protocols.
• This work advances the application of network control theory for targeted interventions in brain function.