Table 2.

nnU-Net Automatic Segmentation Performance. Accuracy represents percentage of overall percentage of correctly predicted areas in the overall image. Sensitivity corresponds to the percentage of nodules that were correctly predicted and annotated. Specificity refers to the percentage of healthy area that was correctly predicted and annotated.

CT Automatic Segmentation Performance
	AUC	Accuracy	Sensitivity	Specificity
Training Set	0.865	0.999	0.730	0.999
Validation Set	0.848	0.999	0.696	0.999
Test Set	0.828	0.999	0.657	0.999

PET Automatic Segmentation Performance
	AUC	Accuracy	Sensitivity	Specificity
Training Set	0.747	0.998	0.495	0.999
Validation Set	0.741	0.999	0.489	0.999
Test Set	0.755	0.999	0.510	0.999