A Comparative Study of Three Systems for Liver Magnetic Resonance Elastography

Area of Science:

• Medical Imaging
• Biophysics
• Radiology

Background:

• Magnetic Resonance Elastography (MRE) systems can yield varying stiffness measurements.
• This variability poses challenges for direct comparison in multi-center research.

Purpose of the Study:

• To evaluate the repeatability and reproducibility of liver stiffness measurements across three distinct MRE systems.
• To assess interobserver, interscan, and intersystem variability in MRE liver stiffness quantification.

Main Methods:

• Prospective study involving 30 healthy volunteers and 5 gel phantoms.
• Utilized three MRE systems: 3.0T UIH, 1.5T Siemens, and 3.0T GE, employing Echo Planar Imaging-based MRE sequences.
• Statistical analysis included Intraclass Correlation Coefficients (ICC), Coefficients of Variation (CV), Bland-Altman analyses, paired t-tests, and Friedman tests.

Main Results:

• High interobserver reproducibility (ICC > 0.92, CV < 4.32%) and excellent interscan repeatability (ICC > 0.96, CV < 3.86%) were observed.
• Intersystem reproducibility for liver evaluations showed good results (ICCs 0.70-0.87, CVs 6.46%-10.99%), with significant differences in absolute stiffness values across systems.
• Phantom studies corroborated these findings, indicating consistent performance across different MRE platforms.

Conclusions:

• Liver stiffness measurements vary between different MRE systems.
• Despite quantitative differences, MRE systems demonstrate excellent reproducibility, supporting their use in multi-center investigations.
• Standardization or calibration may be beneficial for precise cross-system comparisons.