MRI

class deepinv.physics.MRI(mask: Tensor | None = None, img_size: tuple | None = (320, 320), device='cpu', **kwargs)[source]

Bases: DecomposablePhysics

Single-coil accelerated magnetic resonance imaging.

The linear operator operates in 2D slices and is defined as

\[y = SFx\]

where \(S\) applies a mask (subsampling operator), and \(F\) is the 2D discrete Fourier Transform. This operator has a simple singular value decomposition, so it inherits the structure of deepinv.physics.DecomposablePhysics() and thus have a fast pseudo-inverse and prox operators.

The complex images \(x\) and measurements \(y\) should be of size (B, 2, H, W) where the first channel corresponds to the real part and the second channel corresponds to the imaginary part.

A fixed mask can be set at initialisation, or a new mask can be set either at forward (using physics(x, mask=mask)) or using update_parameters.

Note

We provide various random mask generators (e.g. Cartesian undersampling) that can be used directly with this physics. See e.g. deepinv.physics.generator.mri.RandomMaskGenerator

Parameters:
  • mask (torch.Tensor) – binary mask, where 1s represent sampling locations, and 0s otherwise. The mask size can either be (H,W), (C,H,W), or (B,C,H,W) where H, W are the image height and width, C is channels (typically 2) and B is batch size.

  • img_size (tuple) – if mask not specified, blank mask of ones is created using img_size, where img_size can be of any shape specified above. If mask provided, img_size is ignored.

  • device (torch.device) – cpu or gpu.


Examples:

Single-coil accelerated MRI operator with subsampling mask:

>>> from deepinv.physics import MRI
>>> seed = torch.manual_seed(0) # Random seed for reproducibility
>>> x = torch.randn(1, 2, 2, 2) # Define random 2x2 image
>>> mask = 1 - torch.eye(2) # Define subsampling mask
>>> physics = MRI(mask=mask) # Define mask at initialisation
>>> physics(x)
tensor([[[[ 0.0000, -1.4290],
          [ 0.4564, -0.0000]],

         [[ 0.0000,  1.8622],
          [ 0.0603, -0.0000]]]])
>>> physics = MRI(img_size=x.shape) # No subsampling
>>> physics(x)
tensor([[[[ 2.2908, -1.4290],
          [ 0.4564, -0.1814]],

         [[ 0.3744,  1.8622],
          [ 0.0603, -0.6209]]]])
>>> physics.update_parameters(mask=mask) # Update mask on the fly
>>> physics(x)
tensor([[[[ 0.0000, -1.4290],
          [ 0.4564, -0.0000]],

         [[ 0.0000,  1.8622],
          [ 0.0603, -0.0000]]]])
check_mask(mask: Tensor | None = None) None[source]

Updates MRI mask and verifies mask shape to be B,C,H,W.

:param torch.nn.Parameter, float MRI subsampling mask.

update_parameters(mask=None, check_mask=True, **kwargs)[source]

Updates the singular values of the operator.

Examples using MRI:

A tour of forward sensing operators

A tour of forward sensing operators

Self-supervised MRI reconstruction with Artifact2Artifact

Self-supervised MRI reconstruction with Artifact2Artifact

Self-supervised learning with Equivariant Imaging for MRI.

Self-supervised learning with Equivariant Imaging for MRI.