pygrappa.pars

Python implementation of the PARS algorithm.

pygrappa.pars.pars(kx, ky, k, sens, tx=None, ty=None, kernel_size=25, kernel_radius=None, coil_axis=-1)[source]

Parallel MRI with adaptive radius in k‐space.

Parameters:
  • ky (kx,) – Sample points in kspace corresponding to measurements k. kx, kx are 1D arrays.
  • k (array_like) – Complex kspace coil measurements corresponding to points (kx, ky).
  • sens (array_like) – Coil sensitivity maps with shape of desired reconstruction.
  • ty (tx,) – Sample points in kspace defining the grid of ifft2(sens). If None, then tx, ty will be generated from a meshgrid with endpoints [min(kx), max(kx), min(ky), max(ky)].
  • kernel_size (int, optional) – Number of nearest neighbors to use when interpolating kspace.
  • kernel_radius (float, optional) – Raidus in kspace (units same as (kx, ky)) to select neighbors when training kernels.
  • coil_axis (int, optional) – Dimension holding coil data.
Returns:

res – Reconstructed image space on a Cartesian grid with the same shape as sens.
Return type:

array_like

Notes

Implements the algorithm described in [1].

Using kernel_radius seems to perform better than kernel_size.

References

[1]Yeh, Ernest N., et al. “3Parallel magnetic resonance imaging with adaptive radius in k‐space (PARS): Constrained image reconstruction using k‐space locality in radiofrequency coil encoded data.” Magnetic Resonance in Medicine: An Official Journal of the International Society for Magnetic Resonance in Medicine 53.6 (2005): 1383-1392.