Characterizing non-gaussian, high b-value diffusion in liver fibrosis: Stretched exponential and diffusional kurtosis...

Area of Science:

• Biomedical Imaging
• Medical Physics
• Radiology

Background:

• Hepatic fibrosis, a chronic liver disease, is characterized by abnormal molecular diffusion.
• Non-Gaussian diffusion behavior in the liver is a key indicator of fibrotic changes.

Purpose of the Study:

• To investigate non-Gaussian diffusion in a murine liver fibrosis model using advanced MRI models.
• To compare the efficacy of stretched exponential and diffusional kurtosis models in characterizing liver diffusion signal decay.

Main Methods:

• Acquisition of high b-value diffusion MRI data from ex vivo murine liver specimens.
• Analysis of diffusion data using monoexponential, stretched exponential, and diffusional kurtosis models.
• Correlation of diffusion model parameters with liver histopathology.

Main Results:

• Significant correlations were observed between histopathological fibrosis assessments and parameters from stretched exponential and diffusional kurtosis models.
• The diffusional kurtosis model provided a better fit to high b-value diffusion data compared to monoexponential and stretched exponential models, confirmed by Akaike's Information Criterion analysis.

Conclusions:

• Stretched exponential and diffusional kurtosis models can characterize non-Gaussian diffusion in the liver over an extended b-factor range.
• These advanced MRI models offer potential for enhanced characterization of chronic liver disease.