rfDemons: Resting fMRI-based Cortical Surface Registration using the BrainSync Transform.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Brain Mapping

Background:

• Accurate cortical registration is crucial for cross-subject functional brain studies.
• Current methods relying solely on anatomical folding patterns do not precisely align functional regions.
• Existing resting-state fMRI (rfMRI) techniques often use connectivity, which can be indirect.

Purpose of the Study:

• To develop and evaluate a novel cortical surface registration method using synchronized resting-state fMRI data.
• To improve the accuracy of aligning functional brain regions across subjects.
• To assess the impact of rfMRI-based registration on functional alignment.

Main Methods:

• Introduced the rfDemons method for cortical surface registration.
• Utilized synchronized resting-state fMRI time series directly, not connectivity-based features.
• Employed the BrainSync transform for temporal alignment of rfMRI data.
• Initialized registration with curvature-based anatomical registration.

Main Results:

• The rfDemons method was successfully applied to rfMRI data from the Human Connectome Project.
• Evaluation using task fMRI data demonstrated the method's effectiveness in functional alignment.
• The proposed method offers an alternative to connectivity-based approaches for rfMRI registration.

Conclusions:

• Cortical registration based on synchronized rfMRI data improves functional alignment accuracy.
• The rfDemons method provides a robust approach for cross-subject brain analysis.
• This technique enhances the potential of resting-state fMRI for mapping functional brain organization.