Research Article
Combined CNN and Pixel Feature Image for Fatty Liver Ultrasound Image Classification
Table 3
The comparison of the proposed method with other three methods.
| Methods | Filters | Filter size | Accuracy (%) | Specificity | Sensitivity |
| CNNs with 10 layers () [21] | 32 | | 82% | 0.88 | 0.78 | CNNs with 20 layers () [21] | 32 | | 80% | 0.86 | 0.75 | Our CNN without a skip connection | 198 | | 88% | 0.92 | 0.81 | Our CNN with a skip connection | 198 | | 92% | 0.95 | 0.83 |
|
|