Machine learning models and restricted cubic spline were employed to analyze and predict postoperative ischemic stroke in type A aortic dissection patients

BACKGROUD : Ischemic stroke remains a devastating postoperative complication in Type A aortic dissection (TAAD) patients, contributing significantly to elevated mortality rates. Identifying reliable predictors for ischemic stroke risk is crucial for implementing timely clinical interventions. This study endeavors to develop and validate a machine learning-based predictive model for ischemic stroke risk stratification in TAAD patients undergoing surgical treatment. Methods : This retrospective cohort study analyzed 430 TAAD patients who underwent total aortic arch replacement with frozen elephant trunk implantation at Beijing Anzhen Hospital (2015-2021). The cohort was randomly partitioned into training (70%, n=301) and validation (30%, n=129) sets. Feature selection was performed using Boruta algorithm, with variables demonstrating P<0.1 in univariate analysis subsequently incorporated into multivariate logistic regression. Ten machine learning models were evaluated through receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and calibration plots. Model interpretability was enhanced via Shapley Additive Explanations (SHAP), while restricted cubic splines (RCS) elucidated potential non-linear/liner relationships between predictors and result. Results: The GBM model demonstrated superior predictive performance compared to all other models, achieving an area under the curve (AUC) of 0.804 in the validation cohort. SHAP analysis identified the following key predictors of postoperative ischemic stroke: age, history of cerebrovascular disease, cardiopulmonary bypass time(CPBT), intraoperative blood loss volume(IBLV), and preoperative systolic blood pressure(SBP).Furthermore,RCS were independently constructed for each continuous variable to explore variable-outcome relationships. Conclusion: The Gradient Boosting Machine (GBM) model demonstrates the best predictive capacity for postoperative ischemic stroke in TAAD patients, offering clinicians a clinically actionable tool for early postoperative risk stratification and personalized therapeutic optimization.