Evaluating kurtosis-based diffusion MRI tissue models for white matter with fiber ball imaging

Area of Science:

• Neuroimaging
• Biophysics
• Medical Physics

Background:

• Diffusion MRI (dMRI) enables quantification of brain tissue microstructure.
• Tissue models relate biological features (e.g., cell morphology, permeability) to diffusion dynamics.
• Validation of white matter tissue models is crucial for accurate microstructural parameter estimation.

Purpose of the Study:

• To validate and compare three different dMRI tissue models for white matter microstructure quantification.
• To assess model performance against an independent measurement from fiber ball imaging (FBI).

Main Methods:

• Three dMRI tissue models, compatible with diffusional kurtosis imaging (DKI), were evaluated.
• Model predictions for a microstructural parameter were compared to FBI measurements.
• Analysis focused on regions with varying fractional anisotropy (FA) values.

Main Results:

• All three models showed reasonable agreement with FBI for voxels with FA > 0.5.
• The white matter tract integrity (WMTI) model demonstrated significantly better agreement with FBI at lower FA values (< 0.5).
• This suggests superior performance of the WMTI model in complex white matter regions.

Conclusions:

• The WMTI model offers a more robust and broadly applicable approach for dMRI-based microstructural analysis in white matter.
• Accurate microstructural quantification is essential for understanding neurological conditions and treatment efficacy.