Development of rapid and effective risk prediction models for stroke in the Chinese population: a cross-sectional study

Abstract

ObjectivesThe purpose of this study was to use easily obtained and directly observable clinical features to establish predictive models to identify patients at increased risk of stroke.Setting and participantsA total of 46 240 valid records were obtained from 8 research centres and 14 communities in Jiangxi province, China, between February and September 2018.Primary and secondary outcome measuresThe area under the receiver operating characteristic curve (AUC), sensitivity, specificity and accuracy were calculated to test the performance of the five models (logistic regression (LR), random forest (RF), decision tree (DT), extreme gradient boosting (XGBoost) and gradient boosting DT). The calibration curve was used to show calibration performance.ResultsThe results indicated that XGBoost (AUC: 0.924, accuracy: 0.873, sensitivity: 0.776, specificity: 0.916) and RF (AUC: 0.924, accuracy: 0.872, sensitivity: 0.778, specificity: 0.913) demonstrated excellent performance in predicting stroke. Physical inactivity, hypertension, meat-based diet and high salt intake were important prediction features of stroke.ConclusionThe five machine learning models all had good predictive and discriminatory performance for stroke. The performance of RF and XGBoost was slightly better than that of LR, which was easier to interpret and less prone to overfitting. This work provides a rapid and accurate tool for stroke risk assessment, which can help to improve the efficiency of stroke screening medical services and the management of high-risk groups.

Acute ischemic stroke prediction and predictive factors analysis using hematological indicators in elderly hypertensives post-transient ischemic attack

“…The performance of these models was as follows: Logistic Regression (Recall: 0.75, Specificity: 0.68, AUC: 0.72), SVM (Recall: 0.70, Specificity: 0.72, AUC: 0.71), and Random Forest (Recall: 0.62, Specificity: 0.79, AUC: 0.71) 49 . Yuexin Qiu et al compared multiple tree-based models after hyperparameter tuning in a large sample study of 46,240, finding the best performances in random forest (sensitivity: 0.778, specificity: 0.913, AUC: 0.924) and XGBoost (sensitivity: 0.776, specificity: 0.916, AUC: 0.924) 50 . Chuan Hong et al, using neural networks and random survival forests on data from diverse large-scale studies in Western populations, fitted models for subgroups based on race, sex, and age, with the highest AUC for neural networks at 0.75 and for random survival forests at 0.73 51 .…”