Four-dimensional virtual catheter: Noninvasive assessment of intra-aortic hemodynamics in bicuspid aortic valve disease

Background: Four-dimensional (4D) flow MRI enables the evaluation of blood flow alterations in patients with congenital bicuspid aortic valve (BAV). However, current analysis methods are cumbersome and lack the use of the volumetric data from 4D MRI. Purpose: To investigate the feasibility and reproducibility of a technique that uses a catheter-like mathematical model (virtual catheter) to assess volumetric intra-aortic hemodynamics from 4D flow MRI in patients with BAV. Materials and Methods: In this retrospective study, data were collected from adult patients with BAV and healthy participants who underwent aortic 4D flow MRI from November 2011 through August 2014. Reproducibility was tested in healthy study participants who underwent test-retest examinations within 2 weeks. Patients were grouped on the basis of the severity of aortic valve regurgitation (AVR) and aortic valve stenosis (AVS). A 4D virtual catheter mathematical model for probing intra-aortic hemodynamic flow was constructed as a tube with an automatically derived radius along the entire thoracic aorta centerline. Volumetric intra-aortic hemodynamics were computed from 4D flow MRI only within the virtual catheter, and the following volume-normalized systolic peaks were derived: kinetic energy (KE), viscous energy loss rate (VELR), and vorticity. Hemodynamic data were presented as medians with interquartile ranges and compared by using Mann-Whitney U test and Kruskal-Wallis test. Results: The study included 91 participants (57 patients [mean age, 46 years 6 12], 18 women; 34 healthy participants [mean age: 44 years 6 14], 12 women; 15 healthy participants underwent test-retest examinations). Patients showed higher VELR values compared with healthy participants (median, 31 W/m³ [interquartile range, 21-72] vs 23 W/m³ [interquartile range, 17-30], respectively; P , .001) and vorticity (69 sec²¹ [interquartile range, 59-87] vs 60 sec⁻¹ [interquartile range, 50-67], respectively; P , .001). Four-dimensional virtual catheter showed differences among different AVS and AVR grades with the highest VELR (120 W/ m³; interquartile range, 99-166; P , .001) and vorticity (108 sec²¹; interquartile range, 84-151; P , .001) found in severe AVS. High test-retest reproducibility was found for all virtual catheter-derived metrics (intraclass correlation, 0.80 6 0.07; coefficient of variation, 9% 6 3). Conclusion: The proposed four-dimensional (4D) virtual catheter technique enabled reproducible automated evaluation of volumetric intra-aortic hemodynamics alterations from 4D flow MRI in patients with bicuspid aortic valve.