Intra- and Interobserver Variability of Aortic Aneurysm Volume Measurement With Fast CTA Postprocessing Software

Abstract

To study inter- and intraobserver variability of volume measurements with a new, fast semi-automatic method and compare the results to a previously validated standard manual method for volume measurement. Twenty abdominal computed tomographic angiography (CTA) datasets of patients with abdominal aortic aneurysms undergoing endovascular aneurysm repair (EVAR) were randomly selected from a clinical database (10 pre and 10 post EVAR). Aneurysm sac volume was measured by 2 independent observers using both the standard and semi-automatic methods. Intra- and interobserver variabilities and variability between the 2 methods were studied. Differences of each pair of measurements were plotted against their mean, and the repeatability coefficient (RC) was calculated according to Bland and Altman. For the standard method, the intraobserver mean difference was 0.9 mL (0.4% of the first measurement), with an RC of 8.4 mL (4.2%); the interobserver mean difference was 0.0 mL (0.0%), with a RC of 11.8 mL (5.9%). For the semi-automatic method, the intraobserver mean difference was 1.4 mL (0.7%), with an RC of 7.8 mL (4.1%); the interobserver mean difference was -1.8 mL (-1.0%), with an RC of 10.8 mL (5.7%). The mean difference between the methods was 8.3 mL (4.2%), with an RC of 25.1 mL (12.6%) for observer 1, and a mean difference of 6.4 mL (3.2%) and an RC of 21.3 mL (10.7%) for observer 2. The semi-automatic method showed good intra- and interobserver variability for volume measurements of aortic aneurysms before and after EVAR. Volume measurements with the semi-automatic method correspond to measurements with the standard method. Sophisticated and fast postprocessing software may facilitate acceptance and clinical application of volume measurements in daily practice. The cutoff of 5% for relevant volume changes during follow-up, as advised in the reporting standards for EVAR, is sustained.