Research Article
Breast Cancer Pathological Image Classification Based on the Multiscale CNN Squeeze Model
Table 8
Comparison of classification results of each fusion method under different number of feature scales.
| Number of scales | Fusion method | % | % |
| 2 | Add | 86.28 | 86.30 | 2 | Max | 85.65 | 85.97 | 2 | Cat1(sign) | 84.88 | 85.01 | 2 | Cat1 | 86.86 | 86.28 | 2 | Cat2(sign) | 85.29 | 85.26 | 2 | Cat2 | 87.40 | 85.90 | 3 | Add | 85.89 | 85.43 | 3 | Max | 86.44 | 86.87 | 3 | Cat1(sign) | 85.89 | 85.77 | 3 | Cat1 | 86.59 | 86.36 | 3 | Cat2(sign) | 85.69 | 86.54 | 3 | Cat2 | 87.29 | 87.09 |
|
|