Complexity organization of resting-state functional-MRI networks

Area of Science:

• Neuroscience
• Complexity Science
• Biomedical Engineering

Background:

• Functional magnetic resonance imaging (fMRI) is used to study brain activity, focusing on resting-state networks (RSNs).
• RSNs are categorized as intrinsic (e.g., default mode network) or extrinsic (e.g., visual network) based on their temporal correlations.
• Entropy measures offer intra-voxel insights into fMRI signals, complementing inter-voxel correlation analyses.

Purpose of the Study:

• To apply multiscale entropy (MSE) to analyze entropy distribution across RSNs using high-quality fMRI data.
• To evaluate MSE's ability to differentiate between various functional brain networks.
• To investigate the complexity hierarchy between intrinsic and extrinsic RSNs.

Main Methods:

• Utilized the Midnight Scan Club dataset, known for its high sampling rate and quality.
• Applied Multiscale Entropy (MSE) analysis to blood oxygen level-dependent (BOLD) time series data.
• Compared entropy profiles across different temporal scales and RSNs, focusing on infra-slow frequencies.

Main Results:

• Spatial distribution of entropy at infra-slow frequencies (0.005-0.1 Hz) successfully reproduced known RSN parcellations.
• A complexity hierarchy was identified, with intrinsic networks consistently showing higher entropy than extrinsic networks.
• The posterior cerebellum demonstrated high entropy levels, comparable to intrinsic RSNs.

Conclusions:

• MSE is a valuable tool for characterizing neural activity complexity within fMRI data.
• The identified entropy hierarchy provides a novel framework for understanding functional brain organization.
• Findings suggest the posterior cerebellum plays a significant role in complex neural processing.