Research Article
Diagnosis of Cervical Cancer based on Ensemble Deep Learning Network using Colposcopy Images
Table 3
Comparative experiment results of proposed architecture with different models.
| Model name | Accuracy (%) | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | Ref |
| DenseNet-121 | 72.42 | 59.86 | 76.83 | 48.39 | 84.52 | [43] | DenseNet-169 | 69.79 | 65.00 | 71.48 | 44.84 | 85.31 | [43] | Colponet | 81.0 | — | — | — | — | [16] | SVM | 63.27 | 38.46 | 71.85 | 32.43 | 76.87 | [44] | Inception-Resnet-v2 | 69.3 | 66.70 | 70.6 | 47.20 | 84.00 | [45] | CYENET | 92.30 | 92.40 | 96.20 | 92.00 | 95.00 | Present study | VGG19 (TL) | 73.30 | 33.00 | 79.00 | 70.00 | 88.00 | Present study |
|
|