Addressing head motion dependencies for small-world topologies in functional connectomics

Area of Science:

• Neuroscience
• Network Science
• Medical Imaging

Background:

• Functional brain connectome analysis using graph theory advances understanding of brain architecture.
• Head motion in resting-state fMRI complicates accurate assessment of brain connectivity and topological parameters.

Purpose of the Study:

• To comprehensively examine the impact of various motion correction strategies on the relationship between head motion and graph theoretical parameters.
• To evaluate individual-level and group-level motion correction techniques in resting-state fMRI data.

Main Methods:

• Individual-level analysis included head motion regression, scrubbing, and partial correlation.
• Group-level analysis involved regression of motion and mean intrinsic functional connectivity before and after topological parameter calculation.
• Global signal regression (GSR) was assessed for its efficacy in mitigating motion-related artifacts.

Main Results:

• Individual-level motion correction methods failed to fully eliminate motion-related effects on topological parameters.
• Global signal regression (GSR) effectively reduced motion-topology relationships but risked altering connectivity structure and hub distribution.
• Group-level motion correction proved crucial for accurate analysis.

Conclusions:

• No single individual-level method completely removes motion artifacts from graph theoretical measures.
• GSR is a potent strategy for motion artifact reduction but requires cautious application, especially with high-density graphs.
• Findings suggest some observed motion-relationships may represent neural signatures rather than pure artifacts, necessitating further investigation.