Structural analysis of fMRI data: a surface-based framework for multi-subject studies

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Brain Imaging Analysis

Background:

• Functional MRI (fMRI) group analysis traditionally uses voxel-based or surface-based methods.
• Existing structural techniques often rely on volume-based representations, which can be sensitive to inter-subject spatial variability.
• Bridging surface-based and structural analysis frameworks offers potential for improved fMRI data interpretation.

Purpose of the Study:

• To present a novel surface-based structural analysis method for fMRI group studies.
• To extend a previous volume-based structural approach to a surface-based domain.
• To jointly optimize inter-subject matching and activation detection for enhanced fMRI analysis.

Main Methods:

• A multi-scale surface-based representation of individual fMRI activation maps is computed.
• Inter-subject matching and activation detection are performed via Markovian model optimization.
• Non-parametric significance measures are calculated to assess result relevance and control Type I error.

Main Results:

• The surface-based structural analysis demonstrates robustness to inter-subject spatial variability.
• The method achieves relevant results with good specificity and sensitivity compared to standard analyses.
• Advantages over 3D structural analysis are highlighted through comparative results.

Conclusions:

• The proposed surface-based structural analysis method offers a robust and accurate approach for fMRI group studies.
• This framework effectively integrates surface-based processing with structural analysis principles.
• The method shows promise for advancing the field of neuroimaging analysis and brain activation detection.