Comparative MR study of hepatic fat quantification using single-voxel proton spectroscopy, two-point dixon and three-point IDEAL.
Hepatic fat fraction (HFF) was measured in 28 lean/obese humans by single-voxel proton spectroscopy (MRS), a two-point Dixon (2PD) and a three-point iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL) method (3PI). For the lean, obese and total subject groups, the range of HFF measured by MRS was 0.3–3.5% (1.1±1.4%), 0.3–41.5% (11.7±12.1), and 0.3–41.5% (10.1±11.6%), respectively For the same groups, the HFF measured by 2PD was −6.3–2.2% (−2.0±3.7%), −2.4–42.9% (12.9±13.8%), and −6.3–42.9% (10.5±13.7%), respectively, and for 3PI they were 7.9–12.8% (10.1±2.0%), 11.1–49.3% (22.0±12.2%), and 7.9–49.3% (20.0±11.8%), respectively. The HFF measured by MRS was highly correlated with those measured by 2PD (r=0.954, p<0.001) and 3PI (r=0.973, p<0.001). With the MRS data as a reference, the percentages of correct diagnosis of fatty liver with the MRI methods ranged 75–93% for 2PI and 79–89% for 3PI. Our study demonstrates that the apparent HFF measured by the MRI methods can significantly vary depending on the choice of water-fat separation methods and sequences. Such variability may limit the clinical application of the MRI methods, particularly when a diagnosis of early fatty liver needs to be performed. Therefore, protocol-specific establishment of cutoffs for liver fat content may be necessary.