POSTER PRESENTATION Open Access Evaluation of Circulation, Γ, as a quantifying metric in 4D flow MRI

4D flow MRI provide a wonderfully rich set of data for diagnosis and understanding of cardiac disease. Interpreta-tion of this wealth of data can be challenging and it may be desirable to simplify it to make analysis manageable. We have observed interesting rotational behavior of blood in the aorta in a number of cardiac diseases. Circulation (Γ) is a metric used in fluid dynamics to quantify the rota-tional components of flow. This metric has not been reported before for 4D flow MRI and may provide a mea-surement that can be used diagnostically. The purpose of this work is to define the normal range of Γ in the ascend-ing aorta and investigate its reproducibility. Methods Circumferential circulation is calculated as the integral of vorticity (ω) with respect to area within a transverse plane through the aorta, Γ = ∫∫ω dS (Stokes ’ theorem). Γ is a 3D vector, whose through plane component, ΓT, depicts the within plane (circumferential) rotation. ΓT was calculated in the ascending aorta at the level of the pulmonary artery, both at peak systole and during mean systole (-80 to 120 ms about peak). 35 healthy volun-teers were assessed, mean age 36.7 ±17.6 years, range from 8 to 74 years old, 23 male and 12 female. Three volunteers were rescanned to assess repeatability. 4D Flow [1] was analyzed in EnSight (CEI, Apex, NC, USA) to calculate vorticity maps. A slice through the lumen was segmented to include only the blood pool, and a discrete integral with respect to voxel cross-sectional area was calculated in Matlab (Natick, MA, USA). Results Figure 1 demonstrates the segmentation on a velocity and vorticity map. In our group of 35 volunteers mean ±SD of ΓT was 2.71 ±2.78 mm 2/s and 3.07 ±4.51 mm2/s for peak and mean systole respectively. Our three repeated scans have differences and means shown in table 1. They show that the intra-subject variation is less than the inter-subject variation, this supports the repro-ducibility of the method. Figure 2 demonstrates the OCM