Quantitative Assessment of Intrahepatic Lipids Using Fat-Selective Imaging With Spectral-Spatial Excitation and In-/Opposed-Phase Gradient Echo Imaging Techniques Within a Study Population of Extremely Obese Patients

Abstract

The aim of this study was to evaluate the feasibility of 2 established magnetic resonance imaging based techniques to quantify intrahepatic lipids (IHL) within a study population of extremely obese patients by means of a short, wide-bore MR scanner. Fat-selective imaging using a spectral-spatial excitation technique and in-phase/opposed-phase (IN/OP) gradient echo imaging were applied and results were compared. Results for IN/OP technique were corrected for T1- and T2*- relaxation effects. Furthermore, image quality was assessed for both techniques. Differences in regional fat distribution were assessed using parameter maps of voxel-wise calculated IHL. MR examinations of 20 extremely obese patients were included in the study (7 males, 13 females; mean age 40.4 +/- 12.6 years; mean body mass index 46.3 +/- 6.6 kg/m2). IHL, in terms of fat signal fractions, was calculated from simultaneously acquired IN/OP-images using a double-echo gradient echo technique. For correction of transverse relaxation effects an additional multiecho gradient echo sequence was applied in each subject, whereas correction of longitudinal relaxation was performed using literature values for T1 of water and lipid protons in the liver parenchyma. A highly selective spectral-spatial excitation technique with 6 binomial radiofrequency pulses was used for fat-selective imaging. In this case, signal intensity of adjacent subcutaneous adipose (approximately 100% fat) was used as an internal reference for IHL quantification. IN/OP-imaging provided sufficient image quality in all subjects, whereas fat-selective imaging was hampered by insufficient homogeneity of the static magnetic field in 1 of 20 subjects. Hepatic T2* values ranged from 20.1 milliseconds to 42.2 milliseconds. Results for IHL from both techniques were highly correlated with r(s) = 0.915 (P < 0.0001). Mean values for IHL were 16.5% +/- 9.2% and 10.6% +/- 7.3%, for IN/OP and spectral-spatial excitation technique, respectively, showing a slightly lower estimation of IHL by the spectral-spatial excitation method. In the examined cohort of extremely obese subjects a relatively high number of 4 out of 20 cases (20%) were found with uneven distribution of IHLs. The presented data confirm that both methods are reliable tools for quantification of IHL, if inherent drawbacks and limitations are taken into account. Inhomogeneity of the static magnetic field observed in examinations of extremely obese patients limits the use of spectral-spatial excitation, if performed without time-consuming shimming procedures. Necessity to correct for transverse and longitudinal relaxation effects using the IN/OP method requires additional measurements and postprocessing procedures, which might hamper the clinical applicability. Moreover, significant regional differences in IHL may exist in some patients especially if pronounced hepatic steatosis is present.