Table 2.

Results of the external validation using CT data sets from two institutions for each of the 8 training models.

Model	Institution 1 (n = 60)			Institution 2 (n = 40)
	Sensitivity	Specificity	Accuracy	Sensitivity	Specificity	Accuracy
1	88.5	91.2	90.0	90.0	90.0	90.0
2	88.5	91.2	90.0	95.0	80.0	87.5
3	92.3	88.2	90.0	95.0	100.0	97.5
4	80.8	91.2	86.7	95.0	95.0	95.0
5	88.5	94.1	92.0	95.0	95.0	95.0
6	88.5	88.2	88.3	90.0	95.0	92.5
7	88.5	88.2	88.3	90.0	90.0	90.0
8	88.5	91.2	90.0	90.0	90.0	90.0
Average	88.5	91.2	90.0	95.0	100.0	97.5