Construction and validation of a prediction model for rupture of anterior communicating artery aneurysms based on morphological characteristics of CTA radiomics
Construction and validation of a prediction model for rupture of anterior communicating artery aneurysms based on morphological characteristics of CTA radiomics
Objective To explore the clinical value of constructing a model for predicting the rupture of anterior communicating artery aneurysms based on CT angiography (CTA) radiomics features. Methods The medical records of 116 patients with anterior communicating artery aneurysms admitted from 2016 to 2023 were retrospectively analyzed and divided into a training set and a test set at a ratio of 8∶2. Head CTA imaging parameters were collected,radiomics features were extracted using 3D Slicer software,and the radiomics score (Rad score) was calculated. A multivariate logistic regression model was used to analyze the risk factors for aneurysm rupture and construct a prediction model. The predictive ability of the model was evaluated using the ROC curve,and the clinical application value was evaluated using the decision curve. Results The multivariate logistic regression analysis showed that age (OR=0.944;95％ CI 0.897～0.993;P=0.025),the ratio of aneurysm length to parent artery diameter (SR;OR=2.247;95％ CI 1.214～4.15;P=0.016),and the ratio of aneurysm height to neck width (aspect ratio,AR;OR=7.942;95％ CI 1.47～42.925;P=0.010) were independent predictors of aneurysm rupture. Four significant radiomics features (Maximum 2D Diameter Column,Maximum 2D Diameter Row,Surface Volume Ratio,Elongation) were screened out by the Lasso regression model,and the Rad score was obtained through calculation. The ROC curve analysis showed that the area under the curve (AUC) of the combined model based on age,AR,SR,and Rad score was 0.889 (95％ CI 0.821～0.958) in the training set and 0.921 (95％ CI 0.803～0.999) in the test set. The calibration plot showed good predictive accuracy between the actual probability and the predicted probability. The decision curve showed that within the threshold probability range of 37％～65％,the net benefit of the combined model was higher than that of the traditional imaging prediction model. Conclusion The combined model constructed by combining CTA radiomics features with traditional imaging features has a good predictive ability for the rupture of anterior communicating artery aneurysms.
维普
万方数据