Table 2.
Teams | Tasks | Key elements in methods | Teams | Tasks | Key elements in methods |
---|---|---|---|---|---|
GUT | CT, MRI | Two-step CNN, combined with anatomical label configurations. | UOL | MRI | MAS and discrete registration, to adapt the large shape variations. |
KTH | CT, MRI | Multi-view U-Nets combining hierarchical shape prior. | CUHK1 | CT, MRI | 3D fully connected network (FCN) with the gradient flow optimization and Dice loss function. |
SEU | CT | Conventional MAS-based method. | CUHK2 | CT, MRI | Hybrid loss guided FCN. |
UCF | CT, MRI | Multi-object multi-planar CNN with an adaptive fusion method. | UT | CT, MRI | Local probabilistic atlases coupled with a topological graph. |
SIAT | CT, MRI | 3D U-Net network learn multi-modality features. | UB2⁎ | MRI | Multi-scale fully convolutional Dense-Nets. |
UB1⁎ | CT, MRI | Dilated residual networks. | UOE⁎ | CT, MRI | Two-stage concatenated U-Net. |
Teams submitted results after the challenge deadline are indicated using Asterisk (*).