Radiology · 1998 · 28 citations · 0 references
MedicineNuclear MedicineRadiologyDiagnostic ImagingDiagnostic RadiologyUltrasonographyCirrhosisHepatobiliary ImagingLiver ImagingMagnetic Resonance ImagingMedical ImagingNuclear Medicine ImagingChemical Shift ImagingFatty Liver Disease ImagingLiver Lesion DetectionLiver Tumor ImagingLiver UltrasoundMagnetic ResonanceSignal Amplitude VariationSignal Variation
PURPOSE: To evaluate proton fat-water chemical shift fast low-angle shot magnetic resonance (MR) imaging for differentiation of fat-containing hyperechoic liver nodules from hyperechoic liver nodules without a fatty component. MATERIALS AND METHODS: T1-weighted fast low-angle shot fat-water chemical shift gradient-echo MR imaging was performed in 96 patients without cirrhosis with 138 hyperechoic liver nodules. In-phase and opposed-phase breath-hold images were acquired. The percentage of signal intensity variation between in-phase and opposed-phase images and the spleen-to-lesion contrast ratio were used to differentiate liver nodules. RESULTS: Chemical shift MR images showed fat in 15 (11%) hyperechoic nodules (two angiomyolipomas and 13 nodular fatty infiltrations of the liver). The mean percentage of signal intensity variation between in-phase and opposed-phase images was 156% (standard error, 43.5%) in nodules with fat and -0.16% (standard error, 0.96%) in nodules without fat (P = .003). Spleen-to-lesion contrast was similar on in- and opposed-phase images in lesions without fat (mean difference, -0.0107; standard error, 0.012), whereas the mean difference in fat-containing nodules was 0.805 (standard error, 0.225; P = .003). The area under the receiver operating characteristic curve was 0.97 for signal intensity variation. CONCLUSION: Hyperechogenicity in certain liver nodules is caused by fat. Chemical shift MR imaging allows accurate differentiation between these and other hyperechoic lesions with no fat component.