Table 2.

Pairwise comparison of ROC curve for classification of six different types of DISs in the testing dataset.

	Difference between Areas	SE	95% CI	p-Value
Panoramic and periapical images vs. oeriapical images	0.007	0.007	−0.008–0.022	0.365
Panoramic and periapical images vs. panoramic images	0.025	0.021	−0.016–0.067	0.235
Panoramic images vs. oeriapical images	0.032	0.020	−0.006–0.072	0.106

AUC, area under the curve; ROC, receiver operating characteristic curve; SE, standard error; CI, confidence interval; AUCs were compared using DeLong’s method for paired ROC curves; panoramic and periapical images, dataset consisting of 2396 panoramic and periapical radiographic images; panoramic images, dataset consisting of 1429 panoramic radiographic images; periapical images, dataset consisting of 967 periapical radiographic images.