Test-retest reliability of white matter structural brain networks: a multiband diffusion MRI study.

Area of Science:

• Neuroimaging
• Brain Network Analysis
• Diffusion MRI

Background:

• The multiband EPI sequence accelerates MRI data acquisition for the Human Connectome Project.
• Test-retest (TRT) reliability of white matter (WM) structural brain network graph metrics from this sequence is uninvestigated.
• Understanding reliability is crucial for clinical applications and individual difference studies.

Purpose of the Study:

• To assess the TRT reliability of WM structural brain networks derived from multiband dMRI.
• To evaluate the impact of parcellation methods (volume- vs. surface-based) and resolution (low- vs. high-) on network reliability.
• To compare the reliability of multiband dMRI-derived networks with conventional dMRI.

Main Methods:

• Utilized a multiband dMRI dataset with repeated scanning sessions.
• Constructed WM networks using both volume- and surface-based parcellation at low and high resolutions.
• Assessed reproducibility using the intra-class correlation coefficient (ICC).

Main Results:

• WM structural networks from multiband dMRI exhibit conserved topological architecture.
• First-order global network metrics are more reliable than second-order metrics.
• Volume-based parcellation and high-resolution networks show better TRT reliability for global metrics.
• Association cortices and hub regions demonstrate higher reproducibility.
• Multiband dMRI networks show higher reproducibility than conventional dMRI networks.

Conclusions:

• WM structural brain networks derived from the multiband EPI sequence demonstrate fair to good TRT reliability.
• The findings support the utility of multiband dMRI for research on individual differences and clinical applications.
• Network construction parameters significantly influence reliability, with volume-based parcellation and higher resolution generally yielding better results for global metrics.