Controlling the human connectome with spatially diffuse input signals

Area of Science:

• Neuroscience
• Network Science
• Computational Biology

Background:

• The human brain exhibits continuous dynamic activity, transitioning between various brain states.
• Network control theory offers a framework for analyzing the energy costs of these state transitions.
• Traditional models assume independent node inputs, ignoring the brain's spatial continuity and limited stimulation specificity.

Purpose of the Study:

• To adapt network control models to incorporate spatially extended inputs.
• To investigate how realistic input strategies affect the energy required for brain state transitions.
• To identify efficient control strategies and their neurobiological correlates.

Main Methods:

• Adapted network control models to include inputs with influence decaying exponentially with distance.
• Analyzed the impact of spatially extended inputs on energy requirements for state transitions.
• Identified near-optimal control strategies and mapped input site density.

Main Results:

• Spatially extended inputs substantially reduce the energy needed for brain state transitions.
• Near-optimal control strategies significantly decrease the number of required inputs (up to two orders of magnitude).
• Maps of optimal input site density align with independent functional, metabolic, genetic, and neurochemical maps.

Conclusions:

• Incorporating spatially extended inputs provides a more realistic and energy-efficient framework for brain control.
• This approach leverages spatial dependencies in brain connectivity and activity.
• The findings offer a neurobiologically grounded method for understanding and controlling brain dynamics.