Table 5.
Quantitative comparisons of cross-domain performance among DeepPyramid+ and state-of-the-art methods, with top two results shown in italic and bold, respectively
| Modality | MRI | Cataract surgery | ||||
|---|---|---|---|---|---|---|
| Backbone | Network | IoU (%) | Dice (%) | Network | IoU (%) | Dice (%) |
| VGG16 | CPFNet | 40.66 | 54.24 | UNet+ | 26.26 | 35.59 |
| UNet++ | 36.30 | 49.30 | CPFNet | 25.14 | 34.04 | |
| DeepPyramid+ | 44.43 | 59.11 | DeepPyramid+ | 42.93 | 55.10 | |
| ResNet34 | scSENet | 40.23 | 53.19 | BARNet | 20.22 | 29.31 |
| FEDNet | 33.05 | 44.47 | PAANet | 14.01 | 20.48 | |
| DeepPyramid+ | 41.52 | 56.14 | DeepPyramid+ | 32.87 | 43.96 | |
| ResNet50 | UPerNet | 38.45 | 51.78 | UPerNet | 28.40 | 38.10 |
| DeepLabV3+ | 37.55 | 49.62 | DeepLabV3+ | 9.14 | 14.16 | |
| DeepPyramid+ | 38.89 | 53.00 | DeepPyramid+ | 29.76 | 40.54 | |
The best results are shown in bold and the second-best are shown in results italic