Test-retest reliability of effective connectivity in the face perception network

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Computational Neuroscience

Background:

• Dynamic Causal Modeling (DCM) is a powerful tool for inferring effective connectivity in the brain.
• Understanding the reliability of DCM is crucial for its application in neuroscience research and clinical settings.
• Systematic analyses of within-subject stability for fMRI-based DCM are limited.

Purpose of the Study:

• To investigate the test-retest reliability of an fMRI paradigm for DCM analysis.
• To assess the reliability of both BOLD activity and effective connectivity measures within the face perception network.
• To determine the impact of experiment length on the reliability of DCM results.

Main Methods:

• Examined test-retest reliability of face-specific BOLD activity in 25 healthy volunteers across two sessions.
• Assessed the stability of effective connectivity using Bayesian model selection and parameter estimation in DCM.
• Investigated the influence of experiment duration on BOLD activity and DCM reliability.

Main Results:

• Good to excellent reliability was found for BOLD activity in DCM-relevant regions.
• Excellent reliability was observed for negative free energy, and good reliability for parameter estimation in DCM.
• Reliability of BOLD activity and DCM results showed only a slight decrease with shortened experiment duration.

Conclusions:

• The presented fMRI paradigm offers reliable estimates for activation and effective connectivity measures.
• The paradigm is suitable for studying neural mechanisms in both normal cognitive function and psychiatric disorders.
• The findings support the use of this paradigm for clinical applications in psychiatric research.