mri.generators.cartesian#

Cartesian Kspace-Generators.

At each iteration, a column of kspace is yielded. Transpose your data if you want to work row-wise.

class Column2DKspaceGenerator(full_kspace, mask_cols, max_iter=0, mode='current', start_center=True)[source]#

Bases: mri.generators.base.KspaceGeneratorBase

K-Space generator, at each step a new column fills the existing k-space.

Parameters

  • full_kspace (numpy.ndarray) – Complete kspace_data.

  • mask_cols (array_like) – List of the column indices to use for the mask.

  • max_iter (int, optional) – The maximum number of iteration to do, default is the number of column provided.

  • mode ({"line", "current", "memory"}) – If “line” : at step k, the generator yields only the data of k-th column, and the column number associated. If “current”: at step k, the generator yields an array with the shape of full_kspace, but containing only the data of step k, and the mask associated. If “memory”, same as “current”, but all the data that have been previously acquired fills the kspace-array. Default is “current”.

  • start_center (bool, optional) – Should the acquisition defined by mask_cols be reordered to start from the center of kspace and move outward by alternating left and right. (default: True)

_getitem_memory(idx)[source]#
_getitem_current(idx: int)[source]#
_getitem_line(idx: int)[source]#