Table 3.
Quantitative comparisons among the performance of DeepPyramid+ and alternative methods in organ and disease segmentation, with top two results shown in italic and bold, respectively
| Modality | Endometriosis surgery | MRI | OCT | |||||
|---|---|---|---|---|---|---|---|---|
| Backbone | Network | IoU (%) | Dice (%) | IoU (%) | Dice (%) | IoU (%) | Dice (%) | Avg. IoU (%) |
| VGG16 | UNet+ | 51.02 | 64.94 | 72.44 | 82.30 | 51.89 | 64.95 | 58.45 |
| scSENet | 48.95 | 62.95 | 72.31 | 82.23 | 52.12 | 65.18 | 57.79 | |
| FEDNet | 50.38 | 64.16 | 71.03 | 81.82 | 48.70 | 62.31 | 56.70 | |
| CE-Net | 44.40 | 58.48 | 70.84 | 81.27 | 47.33 | 61.10 | 54.19 | |
| CPFNet | 52.82 | 66.09 | 73.28 | 82.80 | 47.90 | 61.17 | 58.00 | |
| UNetPP | 51.02 | 64.83 | 72.49 | 82.32 | 51.58 | 64.60 | 58.36 | |
| DeepPyramid+ | 53.22 | 66.37 | 76.02 | 85.36 | 50.79 | 63.99 | 60.01 | |
| ResNet34 | scSENet | 46.20 | 60.65 | 72.99 | 82.66 | 49.64 | 63.31 | 56.28 |
| FEDNet | 28.19 | 37.85 | 70.24 | 80.61 | 45.34 | 59.79 | 47.92 | |
| CE-Net | 11.02 | 18.94 | 13.76 | 17.51 | 15.68 | 25.53 | 13.49 | |
| CPFNet | 23.07 | 35.54 | 59.95 | 73.26 | 18.29 | 28.99 | 33.77 | |
| DeepPyramid+ | 54.04 | 67.63 | 73.46 | 83.66 | 51.22 | 64.82 | 59.57 | |
| ResNet50 | UPerNet | 48.93 | 62.56 | 72.73 | 82.68 | 46.17 | 60.18 | 55.94 |
| DeepLabV3+ | 43.00 | 56.66 | 70.83 | 80.79 | 44.64 | 58.86 | 52.82 | |
| DeepPyramid+ | 53.11 | 67.12 | 73.93 | 83.97 | 50.99 | 64.63 | 59.34 | |
The best results are shown in bold and the second-best are shown in results italic