Research Article
FFU-Net: Feature Fusion U-Net for Lesion Segmentation of Diabetic Retinopathy
Table 3
Ablation study of the proposed model against U-Net, U-Net V1, U-Net V2, and U-Net-FL on MA and HE.
| | | MA | | | HE | | Methods | SEN | IOU | DICE | SEN | IOU | DICE |
| FFU-Net | 0.5933 | 0.5610 | 0.7188 | 0.7342 | 0.7365 | 0.8450 | U-Net V2 | 0.5508 | 0.5267 | 0.6669 | 0.6936 | 0.6917 | 0.8177 | U-Net V1 | 0.5172 | 0.4891 | 0.6334 | 0.6598 | 0.6562 | 0.7897 | U-Net-FL | 0.4968 | 0.4626 | 0.6255 | 0.6447 | 0.6425 | 0.7797 | U-Net | 0.4810 | 0.4490 | 0.6197 | 0.6366 | 0.6333 | 0.7755 |
|
|