Room: Exhibit Hall | Forum 8
Purpose: In order to ensure accurate and reproducible liver fat quantification across sites, vendors, and imaging modalities (MRI and CT), quantitative phantoms that mimic the signal of liver fat in both MRI and CT images are needed. The overall purpose of this project is to develop a novel phantom that is compatible with MRI and CT and mimics the signals of liver fat measured in both modalities.
Methods: A CT and MR compatible proton density fat fraction (PDFF) phantom was built by mixing peanut oil with an agar-based emulsion. The oil-emulsion volume ratio was adjusted for 12 PDFF levels (0%-100%). The mixture was fine-tuned by infusing a small dose (260Âµl Omnipaque/40ml agar) of iodinated contrast agent and the vials that contain these mixtures were enclosed in a spherical water compartment that mimics the abdominal X-ray attenuation environment. A standard abdominal CT protocol was performed on a MDCT system. Multi-echo 3D SGRE MRI data were acquired at 1.5T and 3.0T.
Results: Both the CT-attenuation image and MRI-PDFF map demonstrated high image quality. High correlation was observed between CT-attenuation and nominal PDFF. The relationship between MR-PDFF and CT-attenuation is similar for the phantom data (slope = -0.57, intercept = 31.74, r2 = 0.99) and in-vivo data (slope = -0.58, intercept = 38.23, r2 = 0.83). High correlations were observed between MR-PDFF and nominal PDFF for both 1.5T and 3.0T.
Conclusion: This novel CT and MR compatible phantom can mimic liver fat in both imaging modalities. High correlation (CT-attenuation/MR-PDFF vs nominal PDFF) can be used as reference for clinical purposes. However, MR-PDFF vs CT-attenuation relationship does not perfectly agree with in-vivo data. Further optimization of iodine concentration during construction is expected to overcome this limitation. In the future, the phantom will be used in a multi-site, multi-vendor, multi-modality study of feasibility and reproducibility.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by the NIH (K24 DK102595, R41-EB025729) and a grant from the University of Wisconsin SEED Research Program with research support from GE Healthcare. Scott Reeder and Diego Hernando are founders of Calimetrix, LLC. Scott Reeder has ownership interests in Elucent Medical, Reveal Pharmaceuticals and Cellectar Biosciences.