A validation of dynamic causal modelling for 7T fMRI

Area of Science:

• Neuroscience
• Magnetic Resonance Imaging
• Cognitive Science

Background:

• Ultra-high field magnetic resonance imaging (7T MRI) is increasingly used in neuroscience.
• The utility of 7T MRI for effective connectivity analysis using Dynamic Causal Modelling (DCM) remains unevaluated.

Purpose of the Study:

• To validate Dynamic Causal Modelling (DCM) for 7T functional MRI (fMRI) data.
• To assess the reproducibility and efficiency of DCM estimates at 7T compared to 3T fMRI.

Main Methods:

• Evaluated DCM for 3T and 7T fMRI data from fist-closing movements.
• Assessed reproducibility of connectivity estimates using the intra-class correlation coefficient (ICC).
• Quantified DCM efficiency by comparing posterior distribution entropy of model parameters.

Main Results:

• High reproducibility of average and condition-specific connectivity estimates was found between 3T and 7T (ICC = 0.862 and 0.936).
• 7T fMRI data yielded significantly lower posterior entropy, indicating more informative and efficient DCM parameter estimates compared to 3T.
• The model accommodated field-dependent BOLD signal parameters and regional vascularization differences.

Conclusions:

• Dynamic Causal Modelling (DCM) reliably infers effective connectivity changes from 7T fMRI data.
• 7T fMRI enhances the efficiency and informativeness of DCM analyses.
• This validates 7T fMRI as a powerful tool for advanced neuroimaging analyses.