Multi-Level Clustering of Dynamic Directional Brain Network Patterns and Their Behavioral Relevance
Gopikrishna Deshpande1,2,3,4,5,6,7,8, Hao Jia9
1Department of Electrical and Computer Engineering, AU MRI Research Center, Auburn University, Auburn, AL, United States.
Frontiers in Neuroscience
|March 3, 2020
Summary
Dynamic effective connectivity (DEC) reveals brain network states related to memory and emotion, outperforming static methods in predicting behavior. This study introduces a novel approach to analyze directional brain communication over time.
Area of Science:
- Neuroscience
- Cognitive Neuroscience
- Brain Imaging
Background:
- Dynamic functional connectivity (DFC) offers insights into brain function beyond static measures.
- Directional interactions, or effective connectivity, are crucial for neural communication.
- Dynamic effective connectivity (DEC) analysis is underdeveloped, lacking reproducible methods and behavioral validation.
Purpose of the Study:
- To develop and validate a method for estimating reproducible dynamic effective connectivity (DEC) from resting-state fMRI.
- To investigate the temporal dynamics and functional relevance of DEC brain network configurations.
- To assess the behavioral relevance of DEC compared to static effective connectivity (SEC).
Main Methods:
- Employed a dynamic multivariate autoregressive (MVAR) model to estimate DEC.
- Validated the MVAR method using simulations.
- Applied adaptive evolutionary clustering (AEC) to DEC matrices from resting-state fMRI data (N=21, N=232) to identify dynamic network states.
Main Results:
- Identified quasi-stable, alternating directional brain network configurations (DEC states) over time.
- Dominant DEC states involved temporal, motor, parietal, occipital, and frontal cortices.
- DEC states were functionally linked to memory, emotion, execution, and language.
- DEC-derived metrics explained more behavioral variance (70 behaviors) than SEC in a larger cohort (N=232).
Conclusions:
- The dynamic MVAR approach enables reproducible estimation of DEC.
- DEC captures temporally dynamic, directional brain interactions relevant to cognition and behavior.
- DEC provides a more comprehensive understanding of brain function and its link to behavior than static connectivity measures.
