Machine learning-based models for prediction of the risk of stroke in coronary artery disease patients receiving...

Area of Science:

• Cardiovascular Medicine
• Medical Informatics
• Machine Learning in Healthcare

Background:

• Coronary artery disease (CAD) patients undergoing revascularization face a significant risk of stroke.
• Accurate prediction of this risk is crucial for timely intervention and improved patient management.

Purpose of the Study:

• To develop and evaluate machine learning models for predicting stroke risk in CAD patients post-revascularization.
• To identify the most effective model for clinical application.

Main Methods:

• A cohort of 5757 CAD patients from the MIMIC-IV database undergoing revascularization was analyzed.
• Data were split into training (n=4029) and testing (n=1728) sets.
• Pearson correlation and LASSO regression were used for feature selection, with model performance assessed using AUC, sensitivity, and specificity.

Main Results:

• The Catboost model achieved the highest predictive performance with an AUC of 0.831 in the training set and 0.760 in the testing set.
• The logistic regression model showed an AUC of 0.789 (training) and 0.731 (testing).
• The Catboost model's predictive value was significantly higher than logistic regression (P<0.05), with Charlson Comorbidity Index identified as a key predictor.

Conclusions:

• The Catboost model is optimal for predicting stroke risk in CAD patients post-revascularization.
• This model can serve as a valuable tool for early identification of high-risk patients, potentially reducing postoperative stroke incidence.