Key result
A clinicoradiomics model integrating ascending aortic PVAT radiomics with clinical predictors outperformed clinical or radiomics models alone for 3-month mortality (validation AUC 0.855, P<0.05).
Why the study?
Whether ascending aortic perivascular adipose tissue (PVAT) radiomic features provide incremental prognostic value beyond conventional clinical predictors for short-term mortality after surgical repair of acute type A aortic dissection (ATAAD) was unknown.
Does integrating ascending aortic perivascular adipose tissue radiomics with clinical predictors improve the prediction of 3-month postoperative mortality in patients undergoing surgical repair for acute type A aortic dissection?
Cohort (n=241)
Yes
Does integrating ascending aortic perivascular adipose tissue radiomics with clinical predictors improve the prediction of 3-month postoperative mortality in patients undergoing surgical repair for acute type A aortic dissection?
Integrating ascending aortic PVAT radiomics from preoperative CTA with clinical variables significantly improves short-term mortality risk prediction after surgical repair of acute type A aortic dissection.
Answers are grounded in this paper and cite their sources.
Should not yet inform postoperative risk stratification after acute type A aortic dissection repair; leaves open incremental value pending prospective validation.
The purpose of this study is to investigate whether radiomic features derived from ascending aortic perivascular adipose tissue (PVAT) provide incremental prognostic value beyond conventional clinical predictors for predicting short-term mortality after surgical repair of acute type A aortic dissection (ATAAD). In this retrospective multicenter study, 241 patients with ATAAD who underwent surgical repair were enrolled and divided into a training cohort (n = 180) and an external validation cohort (n = 61). Radiomic features were extracted from preoperative computed tomography angiography (CTA) images of the ascending aortic PVAT extending from the sinotubular junction to the origin of the brachiocephalic trunk. The best-performing radiomics model was integrated with independent clinical predictors to develop a clinicoradiomics model. Model performance was assessed using receiver operating characteristic analysis, calibration, decision curve analysis, DeLong testing, and SHapley Additive exPlanations (SHAP) analysis. Age, neutrophil count, creatinine level, and maximum ascending aortic diameter were identified as independent predictors of 3-month postoperative mortality. The random forest algorithm achieved the best radiomics performance. The clinicoradiomics model achieved superior discrimination, with AUCs of 0.935 and 0.855 in the training and external validation cohorts, respectively, significantly outperforming both the clinical model and the radiomics model (all P < 0.05). SHAP analysis identified multiple PVAT radiomic features among the most influential predictors. Ascending aortic PVAT radiomics provides complementary prognostic information beyond conventional clinical predictors for short-term mortality after surgical repair of ATAAD. A clinicoradiomics model integrating PVAT radiomics with clinical variables improves postoperative risk prediction and may facilitate individualized risk stratification.
No takes yet. Share an insight, caveat, or question.
Shuai et al. (2026) conducted a cohort in Acute type A aortic dissection (ATAAD) (n=241). Clinicoradiomics model integrating ascending aortic perivascular adipose tissue (PVAT) radiomics from preoperative CTA with clinical predictors vs. Conventional clinical model alone and radiomics model alone was evaluated on 3-month postoperative mortality (AUC 0.935 (training) and AUC 0.855 (validation), p=<0.05). A clinicoradiomics model integrating ascending aortic PVAT radiomics with clinical predictors outperformed clinical or radiomics models alone for 3-month mortality (validation AUC 0.855, P<0.05).
Synapse has enriched 10 related papers. Consider them for comparative context: