Table 3 Comparison of different algorithms on the DRIVE dataset.
From: Enhanced feature dynamic fusion gated UNet for robust retinal vessel segmentation
Methods | Date | ACC | SE | SP | AUC | F1 |
|---|---|---|---|---|---|---|
U-Net1 | 2015 | 0.9691 | 0.7948 | 0.9860 | 0.9845 | 0.8172 |
Attention U-Net17 | 2018 | 0.9661 | 0.8092 | 0.9813 | 0.9776 | 0.8015 |
LadderNet18 | 2019 | 0.9561 | 0.7856 | 0.9810 | 0.9793 | 0.8202 |
ResUNet++ 19 | 2019 | 0.9662 | 0.7856 | 0.9837 | 0.9801 | 0.8015 |
HRNet20 | 2019 | 0.9683 | 0.8152 | 0.9831 | 0.9855 | 0.8171 |
AG-NET21 | 2019 | 0.9692 | 0.8100 | 0.9848 | 0.9856 | – |
UNet3+ 22 | 2020 | 0.9702 | 0.8081 | 0.9860 | 0.9860 | 0.8254 |
RVSeg-Net23 | 2020 | 0.9681 | 0.8107 | 0.9845 | 0.9817 | 0.8267 |
BEFD24 | 2021 | 0.9701 | 0.8215 | 0.9845 | 0.9867 | – |
SA-UNet25 | 2020 | 0.9698 | 0.8212 | 0.9840 | 0.9864 | – |
VSSC-Net26 | 2021 | 0.9627 | 0.7827 | 0.9821 | 0.9789 | – |
SCS-Net27 | 2021 | 0.9697 | 0.8289 | 0.9838 | 0.9837 | – |
MMDC-Net28 | 2022 | 0.9607 | 0.8074 | 0.9755 | 0.9613 | – |
UCR-Net29 | 2022 | 0.9671 | 0.8120 | 0.9822 | 0.9847 | – |
DCA-CNN30 | 2022 | 0.9630 | 0.8745 | 0.9823 | 0.9670 | – |
D-MNet31 | 2022 | 0.9683 | 0.8363 | 0.9811 | – | 0.8211 |
MBSNet32 | 2023 | 0.9692 | 0.8234 | 0.9834 | 0.9873 | 0.8234 |
DCSAU-Net33 | 2023 | 0.9667 | 0.8376 | 0.9792 | 0.9827 | 0.8139 |
SDDC-Net34 | 2023 | 0.9704 | 0.8603 | 0.9808 | 0.9706 | 0.8289 |
BCU-Net35 | 2023 | 0.9667 | 0.8142 | 0.9816 | 0.9791 | 0.8096 |
GDF-Net36 | 2023 | 0.9622 | 0.8291 | 0.9852 | 0.9859 | 0.8302 |
CFFANet37 | 2024 | 0.9560 | – | 0.9760 | – | 0.8290 |
TCDDU-Net38 | 2024 | 0.9698 | 0.8258 | 0.9838 | 0.9868 | 0.8265 |
DMSU-Net +  + 39 | 2025 | – | 0.8374 | 0.9845 | 0.9786 | 0.8275 |
MVM-UNet40 | 2025 | 0.9683 | 0.8547 | 0.9786 | – | – |
EFDG-UNet | 2025 | 0.9736 | 0.8438 | 0.9856 | 0.9886 | 0.8412 |
