Matthew A. Corriere, Arsalla Islam, Timothy E. Craven, Thomas Conlee, Justin Hurie, Matthew Edwards
Wake Forest University School of Medicine, North Carolina, NC.
OBJECTIVES: Although multiple software programs are available for three-dimensional centerline computed tomography angiography reconstruction (3D-CTA), software availability varies and measurement reliability between programs is unknown. We compared measurements from 3D-CTA reconstructions created with three commercially available programs (AquariusNet, Preview and OsiriX) and characterized agreement for endograft component selection for endovascular abdominal aortic aneurysm repair (EVAR).
METHODS: Axial images from 60 CTAs performed before EVAR were reconstructed and measured with each program. Diameters were measured from outer wall to outer wall at sites of planned graft fixation. Lengths were measured from the lowest renal artery to aortic bifurcation and bilateral hypogastric arteries. Agreement was evaluated using plots, intra-class correlation and repeated measures ANOVA. Diameter-based graft component selection agreement based on instructions for use was assessed with weighted kappa statistics.
RESULTS: Mean differences in diameter ranged from 0.18-0.75mm for aortic and 0.09-0.73mm for iliac measurements. Mean differences in centerline length from the lowest renal to hypogastric artery ranged from 3.94-8.55mm on the right and 2.86-8.23mm on the left. Intra-class correlation estimates for aortic diameter, common iliac artery diameter and centerline distances were ≥0.77 for all pair-wise comparisons. Programs also demonstrated similar pair-wise agreement for diameter-based endograft component selection.
CONCLUSIONS: Although differences in mean diameter and centerline length measurements were identified between programs, these differences were small and of unclear clinical significance to EVAR planning. Agreement between measurements and diameter-based endograft component selection are generally similar between programs. Software selection for 3D-CTA reconstruction should be primarily based on cost, availability and operator preference.
AUTHOR DISCLOSURES: T. Conlee: Nothing to disclose; M. A. Corriere: Nothing to disclose; T. E. Craven: Nothing to disclose; M. Edwards: Nothing to disclose; J. Hurie: Nothing to disclose; A. Islam: Nothing to disclose.
Posted April 2013