Reverse engineering the brain input: Network control theory to identify cognitive task-related control nodes.

Area of Science:

• Neuroscience
• Systems Neuroscience
• Computational Neuroscience

Background:

• The human brain processes complex sensory and internal information during cognitive tasks.
• Identifying specific brain inputs driving these tasks is challenging.
• Understanding these inputs is crucial for deciphering brain function.

Purpose of the Study:

• To develop and validate a framework for reverse engineering brain inputs.
• To identify control nodes and their associated inputs during cognitive tasks.
• To apply this framework to real-world neuroimaging data.

Main Methods:

• Developed an input identification framework based on network control theory.
• Validated the framework using synthetic data from a linear system.
• Applied the framework to functional magnetic resonance imaging (fMRI) data from 200 human subjects performing motor tasks.

Main Results:

• The framework accurately reconstructed data and recovered inputs in synthetic tests.
• The model achieved significant neural dynamics reconstruction (EV = 0.779) in motor tasks with sparse inputs.
• Identified 28 control nodes, predominantly located within the established motor system.

Conclusions:

• The proposed framework provides a robust method for identifying brain inputs.
• This tool aids in understanding the control mechanisms and network interactions in the brain.
• The findings offer insights into the neural basis of motor control and cognitive processing.