Sparse DCM for whole-brain effective connectivity from resting-state fMRI data.

Area of Science:

• Neuroscience
• Network Science
• Dynamical Systems
• Computational Neuroscience

Background:

• Contemporary neuroscience utilizes network science and dynamical systems to explore the brain's complex, self-organized structure.
• Understanding directed interactions (effective connectivity) and their role in brain dynamics remains challenging due to limitations in neuroimaging and complex inverse problems.

Purpose of the Study:

• To develop a novel method for estimating whole-brain effective connectivity from resting-state functional magnetic resonance imaging (fMRI) data.
• To address the challenges of solving large-scale inverse problems inherent in estimating brain connectivity from noisy, indirect measurements.

Main Methods:

• Adapted sparse estimation methods within the dynamic causal modeling (DCM) framework.
• Developed a novel model based on a linearized, region-specific haemodynamic response function.
• The algorithm is termed 'sparse DCM'.

Main Results:

• Sparse DCM demonstrated superior performance compared to state-of-the-art methods on both synthetic and real fMRI data.
• Graph-theoretical analysis of estimated whole-brain effective connectivity in healthy individuals revealed asymmetric connectivity patterns.
• Identified distinct roles for brain areas in promoting network segregation versus integration.

Conclusions:

• The novel sparse DCM method offers a robust approach for estimating whole-brain effective connectivity from resting-state fMRI.
• This advancement facilitates a deeper understanding of brain network organization, dynamics, and the functional specialization of brain regions.