Table 2 Comparative discussion of BCDCNN.
From: BCDCNN: breast cancer deep convolutional neural network for breast cancer detection using MRI images
Analysis based upon | Metrics/Methods | DLA-EABA | Ultrafast DCE-MRI | AlexNet | CNN | DL | Gaussian SVM | Proposed BCDCNN |
|---|---|---|---|---|---|---|---|---|
Breast Cancer Patients MRI’s dataset |  |  | ||||||
Training data = 90% | Accuracy (%) | 77.7 | 78.5 | 80.9 | 84.3 | 85.4 | 87.1 | 90.2 |
Sensitivity (%) | 77.8 | 78.6 | 81.0 | 84.0 | 86.0 | 88.1 | 90.6 | |
Specificity (%) | 78.0 | 78.8 | 81.2 | 84.1 | 86.9 | 88.1 | 90.9 | |
K value = 9 | Accuracy (%) | 79.3 | 80.1 | 82.6 | 85.6 | 84.6 | 86.7 | 89.2 |
Sensitivity (%) | 77.0 | 77.8 | 80.2 | 83.5 | 84.9 | 86.8 | 89.4 | |
Specificity (%) | 81.1 | 82.0 | 84.5 | 85.5 | 86.6 | 87.9 | 90.9 | |
Duke-Breast-Cancer-MR dataset | Â | Â | ||||||
Training data = 90% | Accuracy (%) | 82.5 | 83.3 | 85.9 | 86.2 | 87.1 | 87.6 | 89.4 |
Sensitivity (%) | 84.2 | 85.1 | 87.7 | 89.0 | 90.2 | 89.8 | 92.2 | |
Specificity (%) | 77.4 | 78.2 | 80.6 | 83.9 | 85.0 | 86.4 | 92.0 | |
K value = 9 | Accuracy (%) | 79.0 | 79.8 | 82.3 | 85.7 | 86.5 | 86.8 | 88.8 |
Sensitivity (%) | 79.5 | 80.4 | 82.9 | 86.3 | 87.1 | 87.4 | 89.3 | |
Specificity (%) | 82.8 | 83.6 | 86.2 | 86.5 | 87.4 | 87.3 | 90.1 | |
CBIS-DDSM: Breast Cancer Image Dataset | Â | |||||||
Training data = 90% | Accuracy (%) | 76.9 | 77.7 | 80.1 | 83.5 | 84.5 | 86.2 | 89.3 |
Sensitivity (%) | 77.0 | 77.8 | 80.2 | 83.2 | 85.1 | 87.2 | 89.7 | |
Specificity (%) | 77.2 | 78.0 | 80.4 | 83.3 | 86.0 | 87.2 | 90.0 | |
