Classification of Abdominal Aortic Aneurysm Based on Geometric Quantification Measures

An abdominal aortic aneurysm (AAA) is an asymptomatic aortic disease with a survival rate of 20% after rupture. It is a vascular degenerative condition different from occlusive arterial diseases. AAA is a localized enlargement of the abdominal aorta 1.5 times the normal diameter. The size of the aneurysm is the most important determining factor in its clinical management. However, other measures of the AAA geometry that are currently not used clinically may influence its rupture risk. To this end, the goals of this research were to develop an algorithm to calculate the AAA wall thickness and abdominal aortic diameter at planes perpendicular to the vessel centerline, and to quantify the effect of geometric indices derived from this algorithm on the overall classification accuracy of AAA based on their ruptured or unruptured status. Such quantification was performed based on a retrospective review of existing medical records of 150 AAA patients (75 electively repaired and 75 emergently repaired). Using an algorithm implemented within the MATLAB computing environment, the quantification of maximum diameter and wall thickness relative to the AAA centerline revealed that 10 diameter- and wall thickness-related indices had a significant difference in their means compared to calculating the indices relative to the medial axis. Of these 10 indices, 9 were wall thickness-related while the remaining one was the maximum diameter ( Dmax). Dmax calculated with respect to the medial axis is over-estimated for both unruptured and ruptured AAA compared to its counterpart with respect to the centerline. Using a machine learning classification algorithm implemented in RStudio, C5.0 decision trees were used to train a model using 70% of the dataset and testing it on the remaining 30%. The model had average and maximum classification accuracies of 81.0% and 95.5%, respectively, when the 4 diameter-related indices ( Dmax, aneurysm asymmetry, maximum diameter to proximal neck diameter ratio, and bulge length) were calculated with respect to the centerline. The model revealed that the 8 most significant indices in classifying AAA as either ruptured or unruptured are, in order of importance: centerline length of the AAA, L2-norm of the Gaussian curvature, AAA wall surface area, maximum diameter to proximal neck diameter ratio, mean wall thickness variance, average compactness, AAA sac length, and minimum thrombus thickness. Therefore, the conclusion of this work is that the aforementioned 8 geometric indices should be used in a clinical setting to assess the risk of AAA rupture by means of a decision trees classifier model. This work provides support for calculating cross-sectional diameters and wall thicknesses relative to the AAA centerline and using size, wall thickness and surface curvature based indices in classification studies of AAA.