Comparison and Reproducibility of ADC Measurements in Breathhold, Respiratory Triggered, and Free-Breathing Diffusion-Weighted MR Imaging of the Liver

Purpose: To compare and determine the reproducibility of apparent diffusion coefficient (ADC) measurements of the normal liver parenchyma in breathhold, respiratory triggered, and free-breathing diffusion-weighted magnetic resonance imaging (DWI). Materials and Methods: Eleven healthy volunteers underwent three series of DWI. Each DWI series consisted of one breathhold, one respiratory triggered, and two free-breathing (thick and thin slice acquisition) scans of the liver, at b-values of 0 and 500 s/mm(2). ADCs of the liver parenchyma were compared by using nonparametric tests. Reproducibility was assessed by the Bland-Altman method. Results: Mean ADCs (in 10(-3) mm(2)/sec) in respiratory triggered DWI (2.07-2.27) were significantly higher than mean ADCs in breathhold DWI (1.57-1.62), thick slice free-breathing DWI (1.62-1.65), and thin slice free-breathing DWI (1.57-1.66) (P < 0.005). Ranges of mean difference in ADC measurement limits of agreement between two scans were -0.02-0.05 +/- 0.16-0.24 in breathhold DWI, -0.14-0.20 +/- 0.59-0.60 in respiratory triggered DWI, -0.03-0.03 +/- 0.20-0.29 in thick slice free-breathing DWI, and -0.01-0.09 +/- 0.21-0.29 in thin slice free-breathing DWI. Conclusion: ADC measurements of the normal liver parenchyma in respiratory triggered DWI are significantly higher and less reproducible than in breathhold and free-breathing DWI.