Related Experiment Video
Updated: Apr 7, 2026
07:59
Fat-Water Phantoms for Magnetic Resonance Imaging Validation: A Flexible and Scalable Protocol
Published on: September 7, 2018
12.3K
Multi-vendor, multi-field strength validation of MRI-Proton density fat fraction for fat quantification: A
Kuang-Chen Yen1, Po-Chin Liang2, Yueh-Hsun Lu3
1Department of Medical Imaging, National Taiwan University Cancer Center, Taipei, Taiwan; Department of Radiology, National Taiwan University, Taipei, Taiwan.
Summary
Magnetic Resonance Imaging Proton Density Fat Fraction (MRI-PDFF) is accurate and consistent for liver fat measurement across different MRI vendors and field strengths. Minor vendor-specific biases exist, necessitating harmonization for multi-center studies.
Area of Science:
- Medical Imaging
- Quantitative MRI
- Biomarker Validation
Background:
- Liver steatosis quantification is crucial for disease management.
- Magnetic Resonance Imaging Proton Density Fat Fraction (MRI-PDFF) offers a non-invasive method for assessing liver fat.
- Standardization of MRI-PDFF across different platforms is essential for reliable clinical application.
Purpose of the Study:
- To validate the accuracy and consistency of MRI-PDFF measurements.
- To compare MRI-PDFF performance across major MRI vendors (GE, Siemens, Philips).
- To assess performance at different magnetic field strengths (1.5T and 3.0T).
Main Methods:
- A calibrated phantom with known fat fractions (0-40%) was scanned on GE, Siemens, and Philips systems at 1.5T and 3.0T.
- Pearson correlation coefficient (r), Concordance Correlation Coefficient (CCC), and Bland-Altman analysis were used to assess accuracy and agreement.
- Linearly weighted Kappa (κ) statistics evaluated agreement for clinical steatosis grading.
Main Results:
- All systems demonstrated excellent correlation with reference values (r > 0.99).
- Overall agreement (CCC) was high, ranging from 0.9797 to 0.9964.
- Siemens showed no significant bias, while GE and Philips exhibited minor, statistically significant underestimation biases. Agreement for steatosis grading was excellent to perfect (κ = 0.83–1.00).
Conclusions:
- MRI-PDFF is a precise and consistent biomarker for liver fat quantification.
- Minor inter-vendor biases exist due to proprietary algorithms.
- Harmonization of MRI-PDFF is recommended for multi-center clinical trials to ensure consistent results.
