Table 6.
Diagnostic performance of deep convolutional neural network model for differentiating luminal A and non-luminal A subtypes based on tumor sizes in the main cohort.
| Experiment | Model | AUC (%) | ACC (%) | SEN (%) | SPC (%) | YI (%) |
|---|---|---|---|---|---|---|
| Long diameter ≤ 20 mm | EfficientNet-B0 | 80.8 | 75.9 | 78.8 | 68.8 | 47.5 |
| DenseNet-121 | 81.8 | 80.4 | 82.5 | 75.0 | 57.5 | |
| VGGNet-19 | 77.3 | 74.1 | 76.2 | 68.8 | 45.0 | |
| Long diameter > 20 mm | EfficientNet-B0 | 32.6 | 85.4 | 100 | 0 | 0 |
| DenseNet-121 | 43.4 | 21.3 | 7.9 | 100 | 7.9 | |
| VGGNet-19 | 34.1 | 25.8 | 15.8 | 84.6 | 0.4 |
AUC, area under the curve; ACC, accuracy; SEN, sensitivity; SPC, specificity; YI, Youden’s index.