mridc.collections.reconstruction.models.varnet package
Submodules
mridc.collections.reconstruction.models.varnet.vn_block module
- class mridc.collections.reconstruction.models.varnet.vn_block.VarNetBlock(model: torch.nn.modules.module.Module, fft_type: str = 'orthogonal', no_dc: bool = False)[source]
Bases:
torch.nn.modules.module.Module
Model block for end-to-end variational network.
This model applies a combination of soft data consistency with the input model as a regularizer. A series of these blocks can be stacked to form the full variational network.
- forward(pred: torch.Tensor, ref_kspace: torch.Tensor, sens_maps: torch.Tensor, mask: torch.Tensor) torch.Tensor [source]
- Parameters
kspace (Reference k-space data.) –
sens_maps (Coil sensitivity maps.) –
mask (Mask to apply to the data.) –
- Return type
Reconstructed image.
- sens_expand(x: torch.Tensor, sens_maps: torch.Tensor) torch.Tensor [source]
Expand the sensitivity maps to the same size as the input.
- Parameters
x (Input data.) –
sens_maps (Coil Sensitivity maps.) –
- Return type
SENSE reconstruction expanded to the same size as the input sens_maps.
- sens_reduce(x: torch.Tensor, sens_maps: torch.Tensor) torch.Tensor [source]
Reduce the sensitivity maps.
- Parameters
x (Input data.) –
sens_maps (Coil Sensitivity maps.) –
- Return type
SENSE coil-combined reconstruction.
- training: bool