Structural connectome with high angular resolution diffusion imaging MRI: assessing the impact of diffusion weighting...

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Network Science

Background:

• Human brain networks (connectomics) are characterized using topological properties derived from high angular resolution diffusion imaging (HARDI) MRI.
• Computational network analysis provides tools to assess the complex structure of these brain networks.

Purpose of the Study:

• To investigate the effect of varying diffusion weighting (b value) and gradient directions on graph theoretical network metrics in human brain connectomes.
• To compare two distinct HARDI acquisition schemes: low b value/low direction number (LBLD) and high b value/high direction number (HBHD).

Main Methods:

• Probabilistic tractography using Q-ball reconstruction of HARDI MRI data.
• Estimation of structural connections between all pairs of regions defined by the automated anatomical labeling (AAL) atlas.
• Comparison of network metrics between LBLD and HBHD HARDI schemes.

Main Results:

• Both LBLD and HBHD schemes produced highly overlapping hub structures in the connectome.
• The HBHD scheme revealed significantly higher connection probabilities and detected more connections between cortical and subcortical regions.
• The HBHD scheme resulted in reduced small-worldness and modularity but a significantly higher clustering coefficient, indicating increased segregation.

Conclusions:

• The selected HARDI scheme critically influences structural connectome measures, a fact not always evident from tractography alone.
• The number of gradient directions and b values can introduce bias in network property assessment.
• Careful selection and consideration of HARDI protocols are essential for reliable comparisons across connectomic studies.