Staging liver fibrosis by a continuous-time random-walk diffusion model.

Area of Science:

• Medical Imaging
• Hepatology
• Biophysics

Background:

• Noninvasive assessment of liver fibrosis is crucial for clinical management.
• Current noninvasive methods have limitations in accurately staging liver fibrosis.
• Diffusion-weighted imaging (DWI) offers potential for evaluating liver tissue characteristics.

Purpose of the Study:

• To evaluate the clinical potential of a continuous-time random-walk (CTRW) diffusion model for liver fibrosis staging.
• To compare the efficacy of CTRW-derived parameters against conventional noninvasive markers.

Main Methods:

• Prospective study of 52 patients undergoing liver biopsy.
• Multi-b value DWI on a 1.5 T MR scanner to derive CTRW parameters (D, α, β) and apparent diffusion coefficient (ADC).
• Comparison with transient elastography (LSM), APRI, and FIB-4 index against fibrosis stages.

Main Results:

• CTRW parameters (D), ADC, and LSM showed significant correlations with fibrosis stages.
• A combined model incorporating CTRW parameters (D, α, β), ADC, and LSM demonstrated superior diagnostic performance for significant fibrosis (F ≥ 2).
• The combined model outperformed individual parameters and a CTRW-only model for advanced fibrosis (F ≥ 4).

Conclusions:

• A novel model integrating CTRW-derived diffusion parameters (D, α, β), ADC, and LSM enables accurate noninvasive liver fibrosis staging.
• This CTRW-based model represents a reliable tool for noninvasive liver fibrosis evaluation.
• Further validation may establish this model as a standard in clinical practice.