Research Article
NHL Pathological Image Classification Based on Hierarchical Local Information and GoogLeNet-Based Representations
Table 1
Patch level classification accuracy for different feature set.
| Feature Set | CLL | FL | MCL | Total |
| Gist | 0.5424 | 0.5829 | 0.6252 | 0.5850 | LPQ | 0.6556 | 0.8070 | 0.6405 | 0.7083 | | 0.7343 | 0.8171 | 0.7000 | 0.7546 | | 0.7828 | 0.8713 | 0.7414 | 0.8029 | | 0.7394 | 0.8837 | 0.8631 | 0.8348 | | 0.7394 | 0.8837 | 0.8631 | 0.8355 | | 0.8253 | 0.8969 | 0.8324 | 0.8549 | | 0.9283 | 0.9217 | 0.9324 | 0.9271 |
|
|