Table 5. Quantitative evaluation of vessel segmentation algorithms related to the 20 images from the test set.
Method | TPR | FPR | ACC | Improvement () | TPR | FPR | ACC |
Our method | 0.7556 | 0.0344 | 0.9475 | ||||
Jiang et al. [17] | 0.6220 | 0.0318 | 0.9244 | our method via Jiang et al. | 21.5 | 8.18 | 2.7 |
Perez et al. [27] | 0.7123 | 0.0524 | 0.9196 | our method via Perez et al. | 6.1 | −34.35 | 3.0 |
Staal et al. [5] | 0.6969 | 0.0214 | 0.9441 | our method via Staal et al. | 8.4 | 60.75 | 0.4 |
Zana et al. [20] | 0.6125 | 0.0163 | 0.9372 | our method via Zana et al. | 23.4 | 111.04 | 1.1 |
2nd observer | 0.7316 | 0.0208 | 0.9470 | our method via 2nd observer | 3.3 | 65.38 | 0.05 |