Building a predictive model for predictive risk of abdominal aortic aneurysm based on computed tomography imaging and hematology parameters

An abdominal aortic aneurysm (AAA) is a life-threatening condition whose rupture is a significant cause of sudden death. There is currently no comprehensive model that integrates hematological parameters with computed tomography (CT) imaging features. This study aimed to construct and verify a predictive model for AAA rupture using CT imaging and hematological parameters. A retrospective case-control study was conducted in China. It included 230 patients with AAA admitted from January 1, 2021 to April 30, 2025. Patients were grouped into ruptured and unruptured groups based on their AAA condition. A nomogram model was constructed with multivariate logistic regression by screening key variables through LASSO regression. External validation was performed on an additional cohort of 113 patients. The study cohort comprised 230 patients (mean age 63.00 ± 8.23 years, 59.57 % male), including 106 in the ruptured group and 124 in the unruptured group. The baseline characteristics between the training and external validation sets revealed no significant differences in key demographic and physiological variables between the two groups (all P > 0.05). The maximum AAA diameter exceeding 6.520 cm was significantly associated with increased rupture risk. Hematological parameters, specifically RDW >14.825 %, WBC >13.610 × 10^9/L, neutrophil count >12.065 × 10^9/L, and NLR >8.715, were identified as independent predictors of AAA rupture. The AUC of combined ROC for the training set was 0.823, and for the external validation set, it was 0.911. Additionally, a model calibration plot confirmed the accuracy of predicted probabilities compared to observed outcomes, supporting the robustness of our model. This model AUC increased by 0.04 (P = 0.03) compared to the VSG score (AUC = 0.78), and the net reclassification index increased by 23.5 % in small AAA (<5 cm). Combining CT imaging and hematological parameters can effectively predict the risk of AAA rupture.