Table 2 Performance of the baseline model of pre-trained.
From: Soul: An OCTA dataset based on Human Machine Collaborative Annotation Framework
Model | AUC | ACC | G-mean | Kappa | Dice | FDR |
|---|---|---|---|---|---|---|
U-Net | 0.9222 | 0.9058 | 0.7939 | 0.6654 | 0.7213 | 0.1928 |
AttU-Net | 0.9378 | 0.9176 | 0.8281 | 0.7142 | 0.7636 | 0.1763 |
AttResU-Net | 0.9335 | 0.9144 | 0.8123 | 0.6982 | 0.7491 | 0.1651 |
OCTA-NET | 0.9476 | 0.9217 | 0.8196 | 0.7214 | 0.7673 | 0.1284 |
average | 0.9353 | 0.9149 | 0.8135 | 0.6998 | 0.7503 | 0.1657 |
