To evaluate short-term test–retest and interobserver reproducibility of IVIM (intravoxel incoherent motion) diffusion parameters and ADC (apparent diffusion coefficient) of hepatocellular carcinoma (HCC) and liver parenchyma at 3.0T.Materials and Methods:
In this prospective Institutional Review Board (IRB)-approved study, 11 patients were scanned twice using a free-breathing single-shot echo-planar-imaging, diffusion-weighted imaging (DWI) sequence using 4 b values (b = 0, 50, 500, 1000 s/mm2) and IVIM DWI using 16 b values (0–800 s/mm2) at 3.0T. IVIM parameters (D: true diffusion coefficient, D*: pseudodiffusion coefficient, PF: perfusion fraction) and ADC (using 4 b and 16 b) were calculated. Short-term test–retest and interobserver reproducibility of IVIM parameters and ADC were assessed by measuring correlation coefficient, coefficient of variation (CV), and Bland–Altman limits of agreements (BA-LA).Results:
Fifteen HCCs were assessed in 10 patients. Reproducibility of IVIM metrics in HCC was poor for D* and PF (mean CV 60.6% and 37.3%, BA-LA: −161.6% to 135.3% and −66.2% to 101.0%, for D* and PF, respectively), good for D and ADC (CV 19.7% and <16%, BA-LA −57.4% to 36.3% and −38.2 to 34.1%, for D and ADC, respectively). Interobserver reproducibility was on the same order of test–retest reproducibility except for PF in HCC. Reproducibility of diffusion parameters was better in liver parenchyma compared to HCC.Conclusion:
Poor reproducibility of D*/PF and good reproducibility for D/ADC were observed in HCC and liver parenchyma. These findings may have implications for trials using DWI in HCC. J. Magn. Reson. Imaging 2015;41:149–156. © 2014 Wiley Periodicals, Inc.