Table 3 Comparison of the classification (normal, NPs, and IP) performance in the KUAH dataset between deep learning and humans
From: Deep learning model for differentiating nasal cavity masses based on nasal endoscopy images
Class | Recall | Precision | F1-score | AUC for multiclass | Accuracy for each class | |
|---|---|---|---|---|---|---|
Curriculum learning backbone network: InceptionResNetV2 | Normal | 0.90 ± 0.04 | 0.78 ± 0.03 | 0.84 ± 0.03 | 0.95 ± 0.02 | 0.82 ± 0.02 |
NPs | 0.82 ± 0.02 | 0.85 ± 0.01 | 0.84 ± 0.02 | 0.88 ± 0.01 | 0.80 ± 0.02 | |
IP | 0.56 ± 0.04 | 0.81 ± 0.03 | 0.66 ± 0.04 | 0.87 ± 0.03 | 0.85 ± 0.02 | |
Seven otolaryngologists | Normal | 0.98 ± 0.01 | 0.85 ± 0.02 | 0.91 ± 0.01 | 0.95 ± 0.02 | 0.93 ± 0.01 |
NPs | 0.71 ± 0.08 | 0.91 ± 0.02 | 0.80 ± 0.05 | 0.81 ± 0.03 | 0.80 ± 0.04 | |
IP | 0.64 ± 0.08 | 0.48 ± 0.10 | 0.51 ± 0.01 | 0.76 ± 0.04 | 0.86 ± 0.04 | |
Rhinologists | Normal | 0.98 ± 0.02 | 0.87 ± 0.05 | 0.92 ± 0.03 | 0.95 ± 0.02 | 0.94 ± 0.02 |
NPs | 0.75 ± 0.11 | 0.89 ± 0.08 | 0.81 ± 0.04 | 0.81 ± 0.02 | 0.81 ± 0.03 | |
IP | 0.55 ± 0.31 | 0.54 ± 0.31 | 0.46 ± 0.07 | 0.72 ± 0.11 | 0.86 ± 0.05 | |
Senior residents | Normal | 0.98 ± 0.02 | 0.85 ± 0.01 | 0.91 ± 0.01 | 0.95 ± 0.02 | 0.94 ± 0.01 |
NPs | 0.78 ± 0.03 | 0.92 ± 0.01 | 0.84 0.01 | 0.85± 0.11 | 0.84 ± 0.01 | |
IP | 0.67 ± 0.06 | 0.54 ± 0.07 | 0.60 ± 0.01 | 0.80 ± 0.02 | 0.89 ± 0.01 | |
Junior residents | Normal | 0.98 ± 0.03 | 0.83 ± 0.04 | 0.90 ± 0.02 | 0.94 ± 0.02 | 0.93 ± 0.01 |
NPs | 0.62 ± 0.04 | 0.92 ± 0.03 | 0.74 ± 0.02 | 0.78 ± 0.02 | 0.76 ± 0.01 | |
IP | 0.71 ± 0.02 | 0.36 ± 0.02 | 0.47 ± 0.01 | 0.77 ± 0.01 | 0.82 ± 0.01 |