Table 2.

Performance comparison of machine learning models, logistic regression model, and the PHASES score method.

	XGBoost	ANN	SVM	LR	PHASES
Area under the ROC curve	0.881	0.837	0.838	0.801	0.758
Sensitivity	90.9%	74.0%	72.6%	72.0%	79.7%
Specificity	77.0%	78.0%	81.0%	83.0%	64.0%
Balanced accuracy	0.839	0.760	0.765	0.775	0.718