4D Flow MRI in Patients with Asymptomatic Abdominal Aortic Aneurysms: Reproducibility and Clinical Analysis

Abstract

Introduction: Hemodynamics play an as of yet incompletely understood role in the progression (i.e. growth and rupture) of abdominal aortic aneurysms (AAAs), and are therefore relevant to study. Unfortunately, there are few methods to measure in vivo hemodynamics. One method is 4D flow magnetic resonance imaging (MRI), where in vivo blood flow is visualised for the estimation of wall shear stress (WSS). If robust, 4D flow MRI can be a valuable tool in the study of AAA hemodynamics. Therefore, this pilot study evaluated reproducibility of 4D flow MRI and performed a clinical analysis. Methods: Adult patients with an asymptomatic AAA of at least 30 millimetre in diameter were included in this study. 4D flow acquisitions were performed on a 3 Tesla MRI scanner (Philips, Best, Netherlands) in a sagittal oblique volume which ideally covered the abdominal aorta from the celiac artery to both common iliac arteries. Pulse sequence parameters were: spatial resolution = 2.5×2.5×2.5 mm; field of view = 315×275×77.5 mm; temporal resolution = 24 heart phases; velocity encoding = 0.52 metre per second (m/s) in all three velocity encoding directions. Participants were scanned twice with an interval of 1 week. Mimics (Materialise, Leuven, Belgium) was used to segment the aortic lumen between the celiac artery and the aortic bifurcation for the reproducibility analysis, while the aortic lumen inside the AAA (AAA lumen) was segmented for the clinical analysis. In-house MATLAB software was used to measure velocity and to estimate WSS in Pascal (Pa). Reproducibility was tested on a voxel-by-voxel basis by the calculating mean difference and limits of agreement in a Bland-Altman analysis, and with orthogonal regression analysis to calculate the slope, intercept and Pearson's r of the agreement between the two scans. The clinical analysis summarised mean and maximal velocity/WSS, and estimated Pearson's r between maximal AAA lumen diameter and mean velocity/WSS. A P value < .05 was considered significant. Results: Thirteen male participants with a mean maximal AAA diameter of 48 ± 10.4 millimetre (mm) and a mean maximal AAA lumen diameter of 35.92 ± 8 mm were included. Figure 1 shows velocity (A) and WSS (B) for six participants. As shown in table 1, Bland-Altman analysis revealed small differences in velocity and WSS (respectively 0.0004 m/s and 0.0124 Pa), while orthogonal regression showed moderate agreement (slope: 0.99077/0.9885, r = 0.8338/0.7131 for respectively velocity/WSS). Clinical analysis demonstrated a negative correlation between maximal AAA lumen diameter and WSS (r = -0.551, P = .051), but a positive correlation with velocity (r = 0.411, P = .163). Conclusion: 4D flow MRI is a robust method and can therefore be used to study AAA hemodynamics in larger longitudinal studies, with the aim of studying whether velocity and WSS may predict AAA rupture.Table 1Reproducibility and clinical analysisReproducibility in the entire segmentation volume (Celiac artery - aortic bifurcation)Clinical analysis in AAA lumenBland-Altmanorthogonal regression analysismean differencelimits of agreementslopeinterceptPearson´s raveraged mean over 13 participantsaveraged maximal over 13 participantsPearson´s rP valuevelocity (in metre per second)0.0004 ± 0.02890.1230 ± 0.03470.99077 ± 0.18060.0042 ± 0.02560.8338 ± 0.06430.1696 ± 0.03690.3191 ± 0.14010.411.163wall shear stress (in Pascal)0.0124 ± 0.05950.2621 ± 0.1210.9885 ± 0.21050.0487 ± 0.20070.7131 ± 0.16030.2386 ± 0.07900.4733 ± 0.1053-0.551.051 Open table in a new tab Disclosure: Nothing to disclose