On high order tensor-based diffusivity profile estimation.

Area of Science:

• Neuroimaging
• Biophysics
• Medical Physics

Background:

• Diffusion weighted magnetic resonance imaging (dMRI) measures water diffusion in tissues.
• High order tensors (HOTs) model the apparent diffusion coefficient (ADC) profile in dMRI data.
• Existing methods for HOT estimation have limitations.

Purpose of the Study:

• To propose a new method for estimating HOTs from dMRI data using weighted least squares (WLS) optimization.
• To introduce a novel fractional anisotropy (FA) computation from HOTs, avoiding singularities and zeros.
• To evaluate the proposed method against existing techniques.

Main Methods:

• Development of a WLS optimization technique for HOT estimation.
• Formulation of a new singularity-free FA calculation from HOTs.
• Empirical validation using synthetic and real human brain dMRI datasets.

Main Results:

• The proposed WLS method provides more accurate HOT estimation compared to existing approaches.
• The new FA computation method is robust and avoids common artifacts.
• Evaluation demonstrates superior performance on both simulated and real brain data.

Conclusions:

• The novel WLS-based HOT estimation and FA calculation offer improved accuracy and robustness in dMRI analysis.
• This method enhances the reliability of quantitative diffusion MRI measurements.
• The findings have implications for advanced neuroimaging research and clinical applications.