Table 2 Performance indicators of the Resnet50 classification
From: The value of deep neural networks in the pathological classification of thyroid tumors
Classification | Recall | Precision | NPV | Accuracy | Specificity | F1 | AUC | Kappa coefficient |
|---|---|---|---|---|---|---|---|---|
PTC | 98.35% | 89.47% | 98.18% | 93.42% | 88.52% | 93.70% | 0.981 | 0.7713 |
MTC | 33.33% | 87.50% | 94.04% | 93.83% | 99.55% | 48.28% | 0.978 | |
FTC | 66.67% | 76.92% | 97.83% | 96.71% | 98.68% | 71.43% | 0.970 | |
Adenoma | 54.55% | 60.00% | 95.52% | 92.59% | 96.38% | 57.14% | 0.948 | |
Adenomatous goiter | 50.00% | 60.00% | 97.42% | 95.88% | 98.27% | 54.55% | 0,889 | |
Normal thyroid | 100.00% | 88.14% | 100.00% | 97.12% | 96.34% | 93.69% | 0.983 |