Table 2. Results of CADx when using XGBoost and parameter optimization.

Algorithm	Number of trial	Validation loss	AUC	Accuracy
Random	10	0.488	0.838	0.756
Random	100	0.451	0.864	0.771
Random	200	0.440	0.868	0.784
Random	1000	0.422	0.878	0.806
TPE	10	0.494	0.838	0.762
TPE	100	0.427	0.876	0.811
TPE	200	0.419	0.881	0.804
TPE	1000	0.394	0.896	0.820

Abbreviation: computer-aided diagnosis, CADx; support vector machine, SVM; Tree Parzen Estimator, TPE; area under the curve, AUC.