Fig. 3
From: Deep learning model for differentiating nasal cavity masses based on nasal endoscopy images
Results of classification for normal, NPs, and IP: (a) confusion matrix on test datasets (KUAH) and (b) mean receiver operating characteristic (ROC) curve for five-folds. Note: Korea University Anam Hospital (KUAH); nasal polyps (NPs); inverted papilloma (IP). p-values > 0.05 (fold2:0.91; fold3:0.83; fold4:0.14; fold5:0.14) for the best performance among five-folds and other folds and p value < 0.05 (IP: 2.2e-16; NPs: 1.64e-06; Normal: 0.004) for the curriculum learning backbone network: InceptionResNetV2 and CNN