mrinufft.operators#
Collection of operators applying the NUFFT used in a MRI context.
The recommended way to create a Fourier operator is to use the
get_operator() function, select a NUFFT backend and
provide the trajectory, shape and extras parameters. The Fourier operator can
then be used to apply the forward op() or adjoint
NUFFT (adj_op()), or to compute the
pseudo-inverse of the operator (pinv_solver()).
Tip
All the operators abide by the same interface, and could (mostly) be used interchangeably. See MRI-NUFFT Interfaces Convention for a detailed description of the interface.
Functions
Check if a specific backend is available. |
|
Return an MRI Fourier operator interface using the correct backend. |
|
Return a list of backend. |
|
View k-space data as real-valued tensor. |
|
View real-packed k-space tensor as complex. |
|
View image tensor as real channel-packed tensor. |
|
View real channel-packed image tensor as complex. |
Classes
Base Fourier Operator class. |
|
Fourier Operator with B0 Inhomogeneities compensation. |
|
Stacked NUFFT Operator for MRI. |
|
Fourier Operator with subspace projection. |
|
Cartesian Operator for MRI reconstruction. |
|
MRI Transform operator, build around cufinufft. |
|
Stacked NUFFT Operator for MRI using GPU only backend. |
|
BART implementation of MRI NUFFT transform. |
|
MRI operator using ducc0 backend. |
|
MRI Transform Operator using finufft. |
|
Interface for the gpuNUFFT backend. |
|
MRI operator using numpy NUDFT backend. |
|
MRI operator using numpy NUDFT backend. |
|
PyNUFFT implementation of MRI NUFFT transform. |
|
NUFFT using SigPy. |
|
MRI Transform Operator using Tensorflow NUFFT. |
|
MRI Transform Operator using Torch NUFFT for CPU. |
|
MRI Transform Operator using Torch NUFFT for GPU. |
|
Expose an MRINufftAutoGrad as as DeepInv Physics Operator. |