The effect of gradient sampling schemes on diffusion metrics derived from probabilistic analysis and tract-based...

Area of Science:

• Neuroimaging
• Diffusion Tensor Imaging (DTI)

Background:

• Diffusion MRI is crucial for non-invasively mapping white matter architecture.
• Understanding the impact of acquisition parameters on diffusion metrics is essential for reliable brain analysis.

Purpose of the Study:

• To systematically evaluate how diffusion gradient encoding schemes affect fractional anisotropy (FA), mean diffusivity (MD), and the identification of crossing fibers.
• To determine the optimal number of diffusion directions for accurate diffusion metric estimation and fiber tractography.

Main Methods:

• Eight healthy volunteers underwent Spin-Echo Echo-Planar-Imaging with varying numbers of diffusion gradient directions (N(d) from 15 to 127) and two signal averages (NSA).
• FA and MD maps were generated and analyzed using tract-based spatial statistics.
• The number of voxels supporting two fiber populations (NV(2)) was estimated using Bayesian methods.

Main Results:

• Fractional anisotropy (FA) values decreased significantly with increasing N(d) and NSA.
• Mean diffusivity (MD) showed minimal sensitivity to N(d) and NSA.
• The number of voxels identifying crossing fibers (NV(2)) increased significantly with N(d), but not NSA.

Conclusions:

• Accurate estimation of FA and MD relies primarily on signal-to-noise ratio (SNR).
• Differentiating multiple fiber populations necessitates a high diffusion sampling density (more gradient directions).
• Optimizing diffusion encoding schemes is critical for advanced white matter tract analysis.