Table 1.

AUCs of the classifiers for the three different datasets

	Chest/abdomen X-rays	ET tube presence/absence	ET tube low/normal
Untrained AlexNet	0.996 (0.987–1.000)	0.675 (0.537–0.813)	0.741 (0.612–0.870)
Untrained GoogLeNet	0.959 (0.905–1.000)	0.650 (0.505–0.795)	0.748 (0.623–0.873)
Pre-trained AlexNet	1.000 (1.000–1.000)	0.856 (0.758–0.954)	0.791 (0.672–0.910)
Pre-trained GoogLeNet	1.000 (1.000–1.000)	0.989 (0.970–1.000)	0.809 (0.697–0.921)

The numbers in parentheses represent the 95% confidence intervals