Table 6.
Confusion matrix of the classification of superpixels using the optimized Xception network in the Th-MYCN and Th-ALKF1174L /MYCN mouse models in eight categories: region of undifferentiated neuroblasts, necrosis, cluster of lymphocytes (Lym), hemorrhage (blood), empty/white space, muscle tissue and kidney (separate test set of 16 whole-slide images).
| Undifferentiated region | Necrosis | Lym | Differentiation | Blood | Empty space | Muscle | Kidney | |
|---|---|---|---|---|---|---|---|---|
| Undifferentiated region | 1403 | 3 | 0 | 14 | 1 | 0 | 0 | 0 |
| Necrosis | 13 | 1642 | 1 | 26 | 49 | 2 | 5 | 18 |
| Lym | 6 | 5 | 1150 | 0 | 0 | 0 | 0 | 3 |
| Differentiation | 0 | 0 | 0 | 1261 | 0 | 0 | 0 | 0 |
| Blood | 1 | 7 | 0 | 0 | 1327 | 0 | 9 | 0 |
| Empty space | 0 | 2 | 0 | 0 | 0 | 560 | 3 | 2 |
| Muscle | 0 | 2 | 0 | 0 | 1 | 0 | 1176 | 0 |
| Kidney | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1176 |
Overall accuracy = 98.3%, average precision = 98.5%, average recall = 98.4%.
The bold values indicate the correct predictions of the network.