Novel Dynamic Hepatic Magnetic Resonance Imaging Strategy Using Advanced Parallel Acquisition, Rhythmic Breath-Hold Technique, and Gadoxetate Disodium Enhancement.

OBJECTIVE The aim of the study was to evaluate image quality of a dynamic hepatic magnetic resonance (MR) imaging strategy based on advanced parallel acquisition combined with rhythmic breath-hold and gadoxetate disodium enhancement. MATERIALS AND METHODS Twenty-seven patients (21 male/6 female; mean age, 57.3 years) were enrolled in this institutional review board-approved study and underwent MR imaging at 3 T. The sequence (T1 3-dimensional gradient-recalled echo; acceleration factor, 4; reconstruction mode; controlled aliasing in parallel imaging resulting in higher acceleration factors; acquisition time, 10.4 seconds) was repeated at 8 fixed time points within the 3 minutes after contrast agent injection. Image quality was evaluated on a 5-point scale (1, excellent; 5, nondiagnostic). Dynamic sequences were classified according to perfusion phases and contrast characteristics. Artifacts and position of the liver in the z axis were recorded and analyzed. RESULTS Overall image quality was found to be 1.44 (95% confidence interval, 1.18-1.71). Contrast was scored as excellent in 25 of 27 patients for central vessels and 22 of 27 patients for peripheral vessels. Adequate-quality arterial-phase images were obtained in all 27 patients. Double arterial and single arterial phases were acquired in 13 of 27 and 14 of 27 patients (n = 6 arterial dominant, n = 8 early arterial phases), respectively. In 1 (3.7%) of 27 patients, severe respiratory artifacts were seen during an early arterial phase. Artifacts were observed in 21 of 27 patients and rated as mild in 19 of these. Compromised quality was related to receiver coils (17 of 29), parallel imaging (6 of 29), breathing (3 of 29), and other causes (3 of 29). The position of the liver throughout the dynamic phases was highly constant, with a greatest mean shift of +2.9 mm throughout the first dynamic acquisition. CONCLUSIONS Advanced parallel acquisition with rhythmic breath-hold and gadoxetate injection allows arterial phase imaging without breathing artifacts; a decelerated yet normal breathing pattern results in very robust breath-hold depth.