Table 4 Performance indicators for the Densenet121 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 | 93.39% | 96.58% | 93.65% | 96.72% | 96.72% | 94.96% | 0.987 | 0.8146 |
MTC | 80.95% | 77.27% | 98.19% | 96.30% | 97.75% | 79.07% | 0.992 | |
FTC | 73.33% | 73.33% | 98.25% | 96.71% | 98.25% | 73.33% | 0.986 | |
Adenoma | 60.00% | 60.00% | 95.52% | 92.59% | 96.38% | 57.14% | 0.915 | |
Adenomatous goiter | 50.00% | 46.15% | 97.39% | 94.65% | 96.97% | 48.00% | 0.887 | |
Normal thyroid | 100.00% | 89.66% | 100.00% | 97.53% | 96.86% | 94.55% | 0.994 |