Table 1 The influence of the number of iterations of each model on the Accuracy (%)
From: CRANet: a comprehensive residual attention network for intracranial aneurysm image classification
Epoch | ResNet18 | ResNet34 | ResNet50 | ResNet101 |
|---|---|---|---|---|
50 | 94.18 ± 1.3 | 95.55 ± 2.8 | 94.44 ± 0.6 | 94.63 ± 0.7 |
100 | 95.85 ± 1.6 | 95.88 ± 1.6 | 95.57 ± 0.8 | 94.43 ± 0.3 |
150 | 96.20 ± 1.8 | 95.18 ± 1.2 | 96.39 ± 0.6 | 95.26 ± 0.7 |
200 | 96.57 ± 0.9 | 96.10 ± 0.2 | 95.64 ± 0.13 | 95.23 ± 0.5 |
250 | 96.51 ± 0.5 | 95.62 ± 0.3 | 95.91 ± 0.4 | 95.16 ± 0.3 |
300 | 94.18 ± 0.9 | 95.87 ± 0.2 | 95.65 ± 0.2 | 95.49 ± 0.1 |