Table 5.
MRMIL with different backbones | BN, LG, HG classification | Cancer detection | ||
---|---|---|---|---|
Cohen’s Kappa (%) | Acc (%) | AUROC (%) | AP (%) | |
VGG11bn | 81.8 | 92.7 | 98.2 | 97.4 |
VGG13bn | 79.9 | 92.0 | 97.8 | 96.9 |
ResNet34 | 78.7 | 91.6 | 96.9 | 95.3 |
MRMIL with different backbones | BN, LG, HG classification | Cancer detection | ||
---|---|---|---|---|
Cohen’s Kappa (%) | Acc (%) | AUROC (%) | AP (%) | |
VGG11bn | 81.8 | 92.7 | 98.2 | 97.4 |
VGG13bn | 79.9 | 92.0 | 97.8 | 96.9 |
ResNet34 | 78.7 | 91.6 | 96.9 | 95.3 |