Effects of microperfusion in hepatic diffusion weighted imaging.

Area of Science:

• Magnetic Resonance Imaging
• Biophysics
• Medical Imaging

Background:

• Clinical hepatic diffusion weighted imaging (DWI) commonly uses mono-exponential diffusion models.
• These models may not fully capture the complex water molecule movement in liver tissue.
• Microperfusion, the movement of water due to blood flow, can influence diffusion measurements.

Purpose of the Study:

• To demonstrate that mono-exponential diffusion models in the liver are contaminated by microperfusion.
• To establish the necessity of the bi-exponential model for accurate hepatic diffusion assessment.
• To differentiate between molecular diffusion and microperfusion in liver tissue.

Main Methods:

• Diffusion weighted imaging (DWI) was performed on 19 healthy volunteers at 1.5 T.
• Seven b-values were used with fat suppression and respiratory triggering.
• Analysis involved mono-exponential apparent diffusion coefficient (ADC) and bi-exponential models (Dslow, Dfast, and fractions).
• Simulations were conducted using data approximating previous studies' b-values.

Main Results:

• Mono-exponential ADCs varied significantly across five liver regions (P < 0.001).
• Bi-exponential slow diffusion (molecular diffusion) showed no significant regional differences (P = 0.140).
• Bi-exponential fast diffusion (microperfusion) varied significantly (P < 0.001).
• The bi-exponential model provided a significantly better fit to the diffusion data (P < 0.010).

Conclusions:

• Apparent diffusion coefficients in normal liver tissue are location-dependent due to microperfusion.
• The bi-exponential model accurately separates molecular diffusion from microperfusion.
• Bi-exponential modeling is essential for precise hepatic diffusion weighted imaging, avoiding contamination from microperfusion.