Leveraging artificial intelligence to predict ERG gene fusion status in prostate cancer

BMC Cancer

Table 2 Performance metrics of AI-based models at different magnifications

Magnification	AUC	TP	FP	TN	FN	Sensitivity	Specificity	PPV	NPV	Accuracy	F1 score	Cut-off
10 ×	0.82	45	13	58	15	75.0%	81.7%	77.6%	79.5%	78.6%	0.78	0.4
20 ×	0.84	45	12	59	15	75.0%	83.1%	78.9%	79.7%	79.4%	0.79	0.5
40 ×	0.85	45	13	58	15	75.0%	81.7%	77.6%	79.5%	78.6%	0.78	0.35

AUC Area under curve in receiver operator characteristics curve, TP True positive (correctly classified as ERG-positive), FP False positive (incorrectly classified as ERG-positive), TN True negative (correctly classified as ERG-negative), FN False negative (incorrectly classified as ERG-negative), PPV Positive predictive value, NPV Negative predictive value, Cut-off indicates cut-off value of proportion of positive patches that gives best accuracy

ISSN: 1471-2407