Table 2.
The comparison of our model with the previous model.
| model | Number of images train/test/val | Algorithm | ML/DL | discrimination | AUC | SE | SP |
|---|---|---|---|---|---|---|---|
| our model | 169/42/78 | LR | ML | moderate to severe CAG | 0.818 | 0.724 | 0.896 |
| Ma B, 2020 | 534/153/76 | Inception v3 | DL | whether CG | NR | 0.958 | NR |
| Ba W, 2021 | 1008/100/142 | DeepLab v3 | DL | whether CAG | 0.910 | 0.952 | 0.992 |
| Barmpoutis P, 2022 | 85 | GAGL-VTNet | DL | whether CAG | NR | 0.940 | NR |
| Fang S, 2023 | 1745/435/545 | GasMIL | DL | grade of CAG | 0.877 | 0.700 | 0.700 |
CAG, Chronic Atrophic Gastritis; CG, Chronic Gastritis; LR, Logistic Regression; ML, Machine Learning; DL, Deep Learning; AUC, Area Under the Curve; SE, Sensitivity; SP, Specificity; NR, Not Report.