Improved fibre dispersion estimation using b-tensor encoding.

Area of Science:

• Neuroimaging
• Biophysics
• Diffusion Magnetic Resonance Imaging (dMRI)

Background:

• Diffusion MRI (dMRI) is crucial for studying white matter microstructure.
• Estimating fibre dispersion in white matter is challenging due to inherent degeneracy with microscopic diffusion anisotropy.
• Current methods rely on strong assumptions about white matter tissue properties.

Purpose of the Study:

• To present a novel approach for resolving the degeneracy between fibre dispersion and microscopic diffusion anisotropy in dMRI.
• To reduce the reliance on strong biophysical model assumptions for accurate fibre dispersion estimation.

Main Methods:

• Development of a method combining linear (conventional) and spherical tensor diffusion encoding.
• Simulation of multi-compartment data fitted with a single-compartment model to test accuracy under simplified assumptions.
• Validation using in-vivo dMRI data with varying acquisition parameters (b-value, repetition time, echo time, diffusion time).

Main Results:

• The proposed method significantly reduces bias in fibre dispersion estimation (~5x) compared to conventional dMRI techniques, especially under simplified tissue models.
• Consistent fibre dispersion estimates were observed in in-vivo data across various acquisition parameter changes.
• The addition of spherical tensor encoding data markedly decreases sensitivity to model assumptions of tissue microstructure.

Conclusions:

• Combining linear and spherical tensor encoding in dMRI offers a robust solution to the degeneracy problem in fibre dispersion measurement.
• This approach enhances the accuracy and reliability of white matter microstructure analysis.
• The findings suggest improved characterization of white matter integrity and pathology.