The microstructure-weighted human connectome: network properties and structure-function correlations across spatial...

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biophysics

Background:

• Conventional connectome edge weights (e.g., number of streamlines) lack microstructural specificity.
• Biophysical diffusion modeling offers greater detail about white matter microstructure.

Purpose of the Study:

• Investigate if biophysical diffusion model parameters can serve as informative connectome weights.
• Assess the specificity and functional relevance of novel connectome metrics.

Main Methods:

• Used diffusion MRI data from healthy adults.
• Constructed structural networks weighted by intra-axonal signal fraction (f), perpendicular extra-axonal diffusivity (D⊥), number of streamlines (NOS), fractional anisotropy (FA), and radial diffusivity (RD).
• Correlated weighted connectomes with resting-state fMRI and intracranial conduction velocity measurements.

Main Results:

• All weights showed small-world network properties.
• Intra-axonal signal fraction (f), perpendicular extra-axonal diffusivity (D⊥), and normalized NOS captured non-random local organization.
• Only D⊥ demonstrated significant structure-function coupling across all scales and modalities.
• f and radial diffusivity (RD) showed high consistency in regional structure-function coupling.

Conclusions:

• Connectome weights derived from biophysical diffusion modeling, particularly D⊥, capture meaningful aspects of brain network organization.
• These novel metrics offer improved specificity for understanding macroscale brain organization and function.