Cortical Surface-Informed Volumetric Spatial Smoothing of fMRI Data via Graph Signal Processing

Area of Science:

• Neuroimaging
• Signal Processing
• Computational Neuroscience

Background:

• Functional MRI (fMRI) data preprocessing typically involves spatial smoothing.
• Standard volumetric smoothing uses isotropic Gaussian kernels, ignoring brain structure.
• Cortical surface smoothing adapts to morphology but requires data projection.

Purpose of the Study:

• To propose a novel volumetric spatial smoothing method for fMRI data.
• To develop a method that adapts smoothing to underlying brain morphology.
• To improve the analysis of both task-based and resting-state fMRI.

Main Methods:

• Leveraging principles from graph signal processing.
• Developing a volumetric smoothing technique utilizing gray-white and pial cortical surfaces.
• Adapting the filtering process to morphological details at the cortical level.

Main Results:

• The proposed method provides structure-adaptive volumetric smoothing.
• It integrates cortical surface information into volumetric processing.
• Enables more precise analysis by respecting anatomical boundaries.

Conclusions:

• The novel method offers an improved approach to fMRI data preprocessing.
• Structure-adaptive volumetric smoothing enhances the fidelity of fMRI analyses.
• This technique holds potential for more accurate activation mapping and connectivity analysis.