Table 3 Classification performance of XGBoost using clinical features, hematologic biomarkers and their combination, based on the entire cohort.

Feature	Dataset	Accuracy	Sensitivity	Specificity	AUC
Clinical	10fold_CV**	0.827	0.807	0.782	0.830
Clinical	Open test**	0.810	0.830	0.795	0.875
CBC	10fold_CV**	0.682	0.676	0.698	0.700
CBC	Open test**	0.658	0.638	0.658	0.667
BCM	10fold_CV**	0.550	0.563	0.502	0.513
BCM	Open test**	0.590	0.581	0.525	0.544
Clinical + CBC	10fold_CV**	0.836	0.812	0.828	0.844
Clinical + CBC	Open test**	0.840	0. 848	0.810	0.883
Clinical + CBC + BCM	10fold_CV	0.843	0.816	0.831	0.850
Clinical + CBC + BCM	Open test	0.850	0. 850	0.816	0.889

Comparisons of AUCs between the combined features and other feature groups were performed using T-tests based on 1000 permutations. A statistically significant difference is noted: *P-value < 0.05; **P-value < 0.001. CBC Complete blood count, BCM Blood biochemical marker.

Quick links

Search