Research Article
An Intelligent Clinical Decision Support System for Patient-Specific Predictions to Improve Cervical Intraepithelial Neoplasia Detection
Table 10
Diagnostic accuracy of cytology and the PNN for all the cases excluding ASCUS.
| | Cytological diagnosis | PNN diagnosis | | Training set | Validation set | Test set | Training set | Validation set | Test set |
| Histology | | | | | | | Negative/clinically negative | 89.0% | 93.0% | 88.1% | 94.2% | 93.0% | 88.1% | CIN1 | 75.2% | 69.7% | 66.7% | 84.5% | 72.7% | 72.7% | CIN2/3 | 75.6% | 75.0% | 76.2% | 92.7% | 80.0% | 90.5% | SCC/Adeno-Ca | 56.3% | 50.0% | 100.0% | 75.0% | 50.0% | 100.0% | Average accuracy per set | 80.5% | 80.0% | 79.0% | 90.0% | 82.0% | 84.0% | Overall accuracy | 80.2% (481/600 cases) | 87.7% (526/600 cases) |
|
|