Functional information regarding cardiac performance, pressure gradients, and local flow derangement are available from blood acceleration fields. Thus, this study examines a 2D and 3D phase contrast sequence optimized to efficiently encode three-directional, time-resolved acceleration in vitro and in vivo. Stenosis phantom acceleration measurements were compared to acceleration derived from standard velocity encoded phase contrast-magnetic resonance imaging (i.e., "velocity-derived acceleration"). For in vivo analysis, three-directional 2D acceleration maps were compared to velocity-derived acceleration using regions proximal and distal to the aortic valve in six healthy volunteers at 1.5 and 3.0 T (voxel size = 1.4 × 2.1 × 8 mm, temporal resolution = 16-20 ms). In addition, a 4D acceleration sequence was evaluated for feasibility in a healthy volunteer and postrepair biscuspid aortic valve patient with an ascending aortic aneurysm. The phantom magnetic resonance acceleration measurements were more accurate (nonturbulent root mean square error = 2.2 vs. 5.1 m/s(2) for phase contrast-magnetic resonance imaging) and 10 times less noisy (nonturbulent σ = 0.9 vs. 13.6 m/s(2) for phase contrast-magnetic resonance imaging) than velocity-derived acceleration. Acceleration mapping of the left ventricular outflow tract and aortic arch exhibited signal voids colocated with complex flow events such as vortex formation and high order motion. 4D acceleration data, visualized in combination with the velocity data, may provide new insight into complex flow phenomena.
Copyright © 2011 Wiley Periodicals, Inc.