Extended unified SEM approach for modeling event-related fMRI data

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Systems Neuroscience

Background:

• fMRI research increasingly focuses on distributed neural network covariation.
• Event-related fMRI designs pose challenges for modeling dynamic network couplings.
• Existing methods may not fully capture experimental effects on effective connectivity.

Purpose of the Study:

• Introduce the extended unified SEM (euSEM) for analyzing event-related fMRI data.
• Develop a novel approach for generating effective connectivity maps.
• Address limitations in modeling dynamic network changes under experimental manipulation.

Main Methods:

• The extended unified SEM (euSEM) builds upon the unified SEM framework.
• euSEM estimates direct experimental effects on BOLD activity.
• euSEM models the modulatory influence of manipulations on regional couplings.

Main Results:

• Monte Carlo simulations validate euSEM's ability to recover covariance patterns.
• Application to empirical data demonstrates the feasibility of the euSEM approach.
• Comparison with dynamic causal modeling shows comparable results.

Conclusions:

• The euSEM offers a flexible framework for analyzing event-related fMRI data.
• This method can be applied in exploratory, partially exploratory, or confirmatory analyses.
• euSEM enhances the understanding of effective connectivity in neuroscience research.