Stratifying non-alcoholic fatty liver disease using diffusion-weighted imaging-based habitat imaging: A rabbit model...

Area of Science:

• Radiology and Imaging Science
• Hepatology
• Biomedical Engineering

Background:

• Non-alcoholic Fatty Liver Disease (NAFLD) is a spectrum of liver conditions.
• Early detection of Non-alcoholic Steatohepatitis (NASH) is crucial for effective management.
• Current imaging techniques may lack the sensitivity to fully characterize liver tissue heterogeneity in NAFLD.

Purpose of the Study:

• To evaluate diffusion-weighted imaging (DWI)-based habitat imaging for stratifying NAFLD stages.
• To enable early detection of NASH using multiparametric biomarkers.
• To visualize and quantify hepatic tissue characteristics with enhanced precision.

Main Methods:

• A novel habitat imaging framework integrating DWI, intravoxel incoherent motion (IVIM), stretched exponential model (SEM), and diffusion kurtosis imaging (DKI) was applied.
• Four hepatic habitats reflecting perfusion, diffusion heterogeneity, and microstructural complexity were identified.
• Habitat fractions (H1-H4) and heterogeneity (HH) indices were derived using spatial multiparametric pattern recognition.

Main Results:

• NASH was associated with a significantly lower fraction of habitat 1 (H1) and a higher fraction of habitat 4 (H4) (p < 0.05).
• Habitat heterogeneity (HH) was significantly reduced in NASH compared to normal histology and simple steatosis (p < 0.05).
• H1 demonstrated a high area under the curve (AUC) of 0.931 for distinguishing NASH, comparable to other diffusion metrics.

Conclusions:

• DWI-based habitat imaging is a promising tool for NAFLD monitoring.
• This technique effectively detects tissue heterogeneity in the liver.
• The diagnostic performance is comparable to conventional diffusion metrics for identifying NASH.