Table 6.
Comparative analysis of experimental protocols employed by dental disease diagnosis approaches on UFBAUESC dental image data set.
Author, Year (Ref) | Architecture | Task | Dataset Size and Split | Data Augmentation | Hyperparameters | ||||
---|---|---|---|---|---|---|---|---|---|
Train Set | Valid Set | Test Set | Loss Function | Optimizer | Learning Rate | ||||
Yamanakkana et al., 2022 [125] | Two feature aggregation module | Tooth segmentation | 1200 | 150 | 150 | Affine transformation | - | - | - |
Chen et al., 2021 [126] | Multiscale structural similarity | Tooth segmentation root boundary extraction | 1200 | 150 | 150 | - | Custom hybrid loss | Adam | 0.0001 |
Zhao et al., 2020 [127] | Two stage attention segmentation network | Tooth segmentation | 1200 | 150 | 150 | - | Custom hybrid loss | Adam | 0.001 |
Kosh et al., 2019 [128] | Ensemble U-Net | Tooth segmentation | 1200 | - | 300 | Affine transformation | Cross entropy | Adam | 0.0001 |
Silva et al., 2018 [57] | Mask R-CNN | Tooth segmentation | 753 | 452 | 295 | - | - | - | - |