The Influence of Spatial Smoothing Kernel Size on the Whole-brain Dynamic Functional Network Connectivity and...

Area of Science:

• Neuroimaging
• Brain Dynamics
• Functional Connectivity

Background:

• Spatial smoothing is a standard fMRI preprocessing step.
• Previous studies linked static functional connectivity to smoothing kernel size.
• The impact of smoothing on dynamic functional connectivity remains under-investigated.

Purpose of the Study:

• To investigate the effects of spatial smoothing kernel size on dynamic functional connectivity (dFNC) and meta-state analysis.
• To examine how different smoothing widths influence brain dynamics in fMRI data.

Main Methods:

• Applied Gaussian smoothing kernels (4, 8, 12 mm FWHM) during fMRI preprocessing.
• Utilized group independent component analysis (ICA) with a high model order (75).
• Performed dFNC using sliding-time windows and k-means clustering, alongside meta-state analysis involving PCA and ICA.

Main Results:

• Spatial smoothing demonstrated robust effects on connectivity dynamics, particularly for cognitive/attention networks (FDR corrected-p < 0.01).
• Meta-state analyses revealed significant alterations in metrics like the number of meta-states and meta-state changes.
• Differences were most pronounced when comparing 8 mm vs. 12 mm FWHM smoothing.

Conclusions:

• Spatial smoothing kernel size critically influences functional connectivity dynamics and meta-state parameters.
• Findings underscore the importance of assessing preprocessing variance on fMRI analysis outcomes.
• This research provides crucial insights into the impact of smoothing on brain dynamics analysis.