Table 4 Statistical analysis between the proposed method and other methods
From: Deep learning approach for automatic segmentation of ulna and radius in dual-energy X-ray imaging
Comparison with methods | Validation set (Dice, p value) | Testing set (Dice, p value) | Independent testing set (Dice, p value) | |||
|---|---|---|---|---|---|---|
Ulna | Radius | Ulna | Radius | Ulna | Radius | |
U-Net | 0.0072 | 0.0219 | 0.0110 | 0.0032 | 0.0174 | 0.0070 |
FCN | 0.0004 | 0.0006 | 0.0003 | 0.0002 | 7.42 × 10–5 | 0.0005 |
| Â | Validation set (Jaccard, p value) | Testing set (Jaccard, p value) | Independent testing set (Jaccard, p value) | |||
|---|---|---|---|---|---|---|
Ulna | Radius | Ulna | Radius | Ulna | Radius | |
U-Net | 0.0078 | 0.0205 | 0.0102 | 0.0038 | 0.0158 | 0.0057 |
FCN | 0.0003 | 0.0005 | 0.0002 | 0.0002 | 3.78 × 10–5 | 0.0006 |