Measuring Asymmetric Interactions in Resting State Brain Networks.

Area of Science:

• Neuroscience
• Network Science
• Computational Biology

Background:

• Directed graph representations are crucial for understanding brain network influences.
• Existing causality-based methods struggle with functional magnetic resonance imaging (fMRI) data due to temporal resolution limits.
• Time-lag dependent techniques are insufficient for analyzing fMRI signals.

Purpose of the Study:

• To introduce a novel necessity measure for inferring brain region interaction direction and strength.
• To develop partial necessity to differentiate direct and indirect interactions.
• To provide a time-lag independent method suitable for fMRI analysis.

Main Methods:

• Developed a novel necessity measure based on the asymmetry of joint brain activation distributions.
• Extended the measure to partial necessity for distinguishing direct/indirect interactions.
• Applied necessity measures to resting-state fMRI data, using kernel density estimation and Wilcoxon rank-sum tests.

Main Results:

• Necessity measures successfully analyzed resting-state fMRI data, revealing network hierarchy and asymmetry.
• Analysis of the default mode network indicated the posterior cingulate cortex's central role.
• The proposed method is more suitable for fMRI data than traditional time-lag approaches.

Conclusions:

• The novel necessity and partial necessity measures are effective for directed brain network modeling with fMRI.
• These methods overcome limitations of time-lag based approaches in neuroimaging.
• Findings support the posterior cingulate cortex's key role within the default mode network.