Convex Hull-based Quantitative Assessment of Shape Irregularity in Cerebral Aneurysms: Diagnostic Performance of the...

Irregular morphology is an important predictor of rupture risk in cerebral aneurysms. However, its evaluation has largely relied on subjective visual assessment of two-dimensional images, resulting in limited reproducibility. This study aimed to establish a quantitative method for assessing aneurysm shape irregularity and to identify morphological variables that best correspond to expert visual judgment.We retrospectively analyzed 184 aneurysms reconstructed from three-dimensional computed tomography angiography. Two independent neurosurgeons classified aneurysms as irregular shape only when both agreed on irregularity; all others were defined as regular shape. Morphological variables, including conventional geometric parameters and convex hull-based irregularity indices such as the undulation area index, which is derived from the difference in surface area between the aneurysm and its convex hull, were used to quantify deviations from an ideal shape. Receiver operating characteristic curve analysis was performed to evaluate their ability to discriminate irregular shape from regular shape.Interrater agreement was moderate (κ = 0.36). All irregularity indices and most conventional variables were significantly higher in the irregular shape group (all p < 0.001). The receiver operating characteristic analysis showed that undulation area index had the highest diagnostic performance (area under the curve = 0.818, sensitivity = 0.779, specificity = 0.738, and optimal cutoff = 0.013).Among the evaluated morphological variables, undulation area index demonstrated the highest ability to discriminate visually defined irregular shape aneurysms. Undulation area index provides an objective and reproducible measure of aneurysm shape irregularity and may contribute to improved risk assessment.