Table II.

ANN Model prediction results in training cross-validation and the three external validation scenarios (A, B and C) in terms of area under the curve (AUC), accuracy (Acc), true and false positive rate (TPR and FPR). Performance on the training set summarizes predictions of the 10×10-fold CV loops for the model based on no-harmonized features and the one based on harmonized features.

	Cross-validation		External validation
	no-harmonized features	harmonized features	scenario A	scenario B	scenario C
AUC (95% CI)	0.89 (0.83–0.95)	0.90 (0.84–0.96)	0.82 (0.73–0.92)	0.82 (0.73–0.92)	0.83 (0.74–0.92)
Acc [%]	83.2	83.4	76.4	72.2	76.4
FPR [%]	14.9	15.8	29.0	22.6	22.6
TPR [%]	81.4	82.8	80.5	68.3	75.6