Table 2.

Performance of the AI system in the test dataset.

Diagnosis	Sensitivity	Specificity	AUROC	F1 score	Brier score	MAE
Covid-19 pneumonia (n = 20)	80%	80%	0.84	0.73	0.16	0.28
Non-Covid-19 pneumonia (n = 20)	60%	90%	0.88	0.67	0.14	0.26
Other (n = 20)	65%	83%	0.86	0.65	0.15	0.26

AI: artificial intelligence, AUROC: area under the receiver operating characteristics, MAE: mean absolute error.