Intravoxel incoherent motion diffusion-weighted imaging in nonalcoholic fatty liver disease: a 3.0-T MR study.

PURPOSE: To compare pure molecular diffusion, D, perfusion-related diffusion, D*, and perfusion fraction, f, determined from diffusion-weighted (DW) magnetic resonance (MR) imaging on the basis of the intravoxel incoherent motion (IVIM) theory in patients with type 2 diabetes with and without liver steatosis. MATERIALS AND METHODS: This prospective study was approved by the appropriate ethics committee, and written informed consent was obtained from all patients. Between December 2009 and September 2011, 108 patients with type 2 diabetes (51 men, 57 women; mean age, 50 years) underwent 3.0-T single-voxel point-resolved proton MR spectroscopy of the liver (segment VII) to calculate the liver fat fraction from water (4.76 ppm) and methylene (1.33 ppm) peaks, corrected for T1 and T2 decay. Steatosis was defined as a liver fat fraction of at least 5.56%. DW imaging was performed by using a single-shot echo-planar sequence with 11 b values (0, 5, 15, 25, 35, 50, 100, 200, 400, 600, 800 sec/mm2). Liver D, D*, and f were measured and compared in patients with and patients without steatosis (Mann-Whitney test). RESULTS: The mean liver fat fraction was 7.8% (standard deviation, 9%; range, 0.99%-45%). Forty patients had liver steatosis. D was significantly lower in steatotic compared with nonsteatotic livers (mean, 1.03×10(-3) mm2/sec±0.23 [standard deviation] vs 1.24×10(-3) mm2/sec±0.15, respectively; P<.0001), as was D* (mean, 72.2×10(-3) mm2/sec±61.4 vs 110.6×10(-3) mm2/sec±79; P=.0025). However, f was significantly higher in steatotic compared with nonsteatotic livers (mean, 33.8%±9.4 vs 26.9%±8.8; P=.0003). CONCLUSION: D is significantly decreased in steatosis. The reduction in D* reflects decreased liver parenchymal perfusion in steatosis. Therefore, steatosis can affect diffusion parameters obtained with IVIM.

Macroscopic heterogeneity of liver fat: an MR-based study in type-2 diabetic patients.

OBJECTIVE: To assess the heterogeneity of liver fat deposition with MR of the liver in type-2 diabetic (T2D) patients. METHODS: We enrolled 121 consecutive T2D patients. The reference standard was 3.0-T (1)H-MR spectroscopy. Hepatic steatosis was defined as liver fat content (LFC) ≥5.56 %. A triple-echo gradient-echo sequence corrected for T1 recovery and T2* decay was used to calculate LFC in left and right livers and hepatic segments. Analyses were performed using a linear mixed model. RESULTS: Fifty-nine (48.8 %) patients had liver steatosis, whereas 62 (51.2 %) did not. Steatosis was greater in the right than in the left liver (P < 0.0001) [mean difference: 1.32 % (range: 0.01-8.75 %)]. In seven patients (5.8 %), LFC was <5.56 % in one side of the liver, whereas it was ≥5.56 % in the other. Steatosis of the left and right liver was heterogeneous at the segmental level in both non-steatotic (P < 0.001 and P < 0.0001 respectively) and steatotic (P < 0.0001 and P = 0.0002 respectively) patients [mean maximum difference: 3.98 % (range: 0.74-19.32 %)]. In 23 patients (19 %), LFC was <5.56 % in one segment, whereas it was ≥5.56 % in at least one other. CONCLUSION: Overall, the mean segmental/lobar variability of steatosis is low. However, segmental variability can sometimes lead to a misdiagnosis. KEY POINTS: There is a need for methods quantifying steatosis over a large region. Steatosis is usually greater in the right than left lobe of the liver. Steatosis within both left and right hepatic lobes is segmentally heterogeneous. Segmental variability of steatosis can result in misdiagnosis.