Predictive value of machine learning algorithm of coronary artery calcium score and clinical factors for obstructive coronary artery disease in hypertensive patients

Figures

• 

  Fig. 1. Flow chart of the study CAD, coronary artery disease; CCTA, coronary computed tomography angiography; CACS, coronary artery calcium score

• 

  Fig. 2. The area under the curve as a measure of individual model performance for the prediction of obstructive coronary artery disease on coronary computed tomography angiography in training cohort (A) and validation cohort (B) AUC, area under the curve; LR, Logistic Regression; XGBoost, Extreme Gradient Boosting; RF, Random Forest; SVM, Support Vector Machine; NNET, Neural Network

• 

  Fig. 3. The receiver operating characteristic curve from applying the best-performing classifier (XGBoost) built in validation cohort AUC, the area under the curve; XGBoost, Extreme Gradient Boosting

• 

  Fig. 4. Calibration curve for different models for prediction of the likelihood of obstructive CAD LR, Logistic Regression; XGBoost, Extreme Gradient Boosting; RF, Random Forest; SVM, Support Vector Machine; NNET, Neural Network; CAD, coronary artery disease

• 

  Fig. 5. Feature importance plot in the XGBoost model The top 15 clinical variables are shown. The yellow and purple points in each row represent participants having low to high values of the specific variable, while the x-axis gives the SHAP value which affects the model [i.e. does it tend to drive the predictions towards the event (positive value of SHAP) or non-event (negative value of SHAP)] CACS, coronary artery calcium score; XGBoost, extreme gradient boosting; SHAP, Shapley additive explanation values; E/e’, early diastolic transmitral velocity to early mitral annulus diastolic velocity ratio; Carotid artery IMT, Carotid artery intima-media thickness