Table 1.

The confusion matrix of a modified U-Net CNN whose task was to classify 300 chest X-rays with COVID-19 in the test set as positive (pos.) and 300 X-rays from healthy patients as negative (neg.).

Predicted$\backslash$True class	Pos.	Neg.
Pos.	TP = 261	FP = 107
Neg.	FN = 39	TN = 193

The resulting values of different evaluation metrics can be found in Table 4.