Research Article
A Soft Label Method for Medical Image Segmentation with Multirater Annotations
Table 4
Quantitative results with different strategies on the QUBIQ-kidney and prostate test sets.
| Methods | Kidney | Prostrate | #Parameters | | | | |
| MV-UNet | 66.59 | 57.83 | 83.50 | 73.71 | 22.0M | STAPLE-UNet | 65.01 | 56.31 | 83.36 | 73.69 | 22.0M | Average-UNet | 69.33 | 58.21 | 85.82 | 77.02 | 22.0M | SVLS-UNet | 70.04 | 58.65 | 86.11 | 77.38 | 22.0M | LNL | 68.40 | 58.59 | 85.44 | 76.91 | 22.2M | MRNet | 71.36 | 60.43 | 87.39 | 78.14 | 81.1M | Ours | 70.25 | 59.08 | 87.67 | 78.55 | 22.0M |
|
|
The best results are highlighted, and the second best results are italic.
|