Channel Capacity for Time-Resolved Effective Connectivity in Functional Neuroimaging

Area of Science:

• Neuroscience
• Systems Neuroscience
• Computational Neuroscience

Background:

• Understanding temporal brain region influence is key in neuroscience.
• Current effective connectivity methods face challenges in interpretability, scalability, and time-resolved estimation.

Purpose of the Study:

• Introduce information channel capacity (ICC) for measuring directed information transfer between brain regions.
• Combine ICC with a sliding-window approach for time-varying directional interaction estimation.
• Validate ICC across human and rodent neuroimaging datasets for sensitivity, specificity, and temporal variability.

Main Methods:

• Utilized information channel capacity, a model-based measure of directed information transfer.
• Employed a sliding-window framework to estimate dynamic directional interactions.
• Validated the method using human fMRI, rat LFP-fMRI, and mouse Ca2+-fMRI datasets.

Main Results:

• Human fMRI demonstrated ICC's sensitivity to task-related increases in directed interactions within motor regions.
• Rat LFP-fMRI confirmed ICC's specificity, showing minimal spurious directional asymmetry.
• Mouse Ca2+-fMRI revealed ICC's ability to identify reproducible dynamic connectivity states and transitions.

Conclusions:

• Information channel capacity provides a physiologically grounded framework for dynamic directional interaction measurement.
• The method is validated across species and neuroimaging modalities, overcoming limitations of existing approaches.
• ICC enables robust, time-resolved analysis of brain network dynamics.