Table 3.
Statistical tests for the 5 AI models and MedSeg against MD 1 and MD 2.
| Paired t-Test | Mann-Whitney | Wilcoxon | |
|---|---|---|---|
| PSPNet vs. MD 1 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| PSPNet vs. MD 2 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| VGG-SegNet vs. MD 1 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| VGG-SegNet vs. MD 2 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| ResNet-SegNet vs. MD 1 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| ResNet-SegNet vs. MD 2 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| VGG-UNet vs. MD 1 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| VGG-UNet vs. MD 2 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| ResNet-UNet vs. MD 1 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| ResNet-UNet vs. MD 2 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| MedSeg vs. MD 1 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| MedSeg vs. MD 2 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| MD 1 vs. MD 2 | p < 0.0001 | p < 0.0001 | p < 0.0001 |