Table 1.
Sensitivity and specificity of simple algorithm and ANN. OSCC: Oral Squamous Cell Carcinoma, HGD: High Grade Dysplasia, LGD: Low Grade Dysplasia, TP: True positive, TN: True negative, FN: False negative, FP: False positive.
| Method | Diagnosis | Sensitivity (TP/(TP + FN)) | Specificity (TN/(TN + FP)) | PPV | NPV |
|---|---|---|---|---|---|
| Algorithm-Score | OSCC Vs Dysplasia/Benign/Normal | 93(51/55) | 74(87/117) | 63 | 96 |
| Dysplasia VsBenign/Normal | 95(91/96) | 76(16/21) | 95 | 76 | |
| DensNet-201-SVM | OSCC Vs Dysplasia/Benign/Normal | 86(43/50) | 81(179/221) | 51 | 96 |
| Dysplasia Vs Benign/Normal | 84(81/97) | 82(101/124) | 78 | 86 | |
| Inception-ResNet-v2-SVM | HGD Vs LGD/Benign/Normal | 83(63/75) | 69(100/146) | 58 | 89 |