TABLE 3.
Aneurysm rupture risk estimation performance of our method and two related works based on CADA dataset.
Classifier | Method | F2-score | ACC | AUC | Precision | Recall |
XGBoost | Liu et al., 2021 | 0.673 | 0.652 | n/a | 0.583 | 0.700 |
Ours | 0.715 | 0.767 | 0.803 | 0.773 | 0.714 | |
KNN | Ivantsits et al., 2021 | 0.690 | 0.660 | n/a | n/a | n/a |
Ours | 0.789 | 0.791 | 0.811 | 0.755 | 0.803 | |
RF | Ivantsits et al., 2021 | n/a | 0.690 | n/a | n/a | n/a |
Ours | 0.675 | 0.751 | 0.810 | 0.771 | 0.660 | |
SVM | Ours | 0.724 | 0.775 | 0.820 | 0.779 | 0.732 |
LR | Ours | 0.731 | 0.776 | 0.834 | 0.761 | 0.732 |
Better results for each specified classifier are highlighted in bold black. The best results for each of these metrics are highlighted in bold red.