get_grappa_caipi_positions#

mrinufft.trajectories.tools.get_grappa_caipi_positions(img_size: tuple[int, int], grappa_factors: tuple[int, int], caipi_delta: int = 0, acs_region: tuple[int, int] | None = None) → ndarray[tuple[int, ...], dtype[_ScalarType_co]] | tuple[ndarray[tuple[int, ...], dtype[_ScalarType_co]], ndarray[tuple[int, ...], dtype[_ScalarType_co]]][source]#

Generate a Cartesian k-space sampling mask for GRAPPA with optional CAIPI shifts.

This function computes the k-space sampling positions for a GRAPPA (GeneRalized Autocalibrating Partial Parallel Acquisition) pattern, optionally incorporating CAIPI (Controlled Aliasing in Parallel Imaging) shifts. The sampling points are distributed over a Cartesian grid based on the specified GRAPPA acceleration factors and image size.

Parameters:

  • img_size (array_like of int, shape (2,)) – The size of the k-space grid (number of samples) along each dimension, typically corresponding to the phase-encoding and frequency-encoding directions.

  • grappa_factors (array_like of int, shape (2,)) – The GRAPPA acceleration factors along each axis. For example, a factor of [2, 1] means every second line is sampled along the first axis, while every line is sampled along the second axis.

  • caipi_delta (float, optional) – The CAIPI phase shift (in units of k-space fraction) to apply along the second dimension. A nonzero value introduces controlled aliasing between slices or coil elements. Default is 0 (no shift).

  • acs_region (array_like of int, shape (2,)) – The size of the Auto-Calibration Signal (ACS) region along each dimension. This region is fully sampled to allow for accurate GRAPPA kernel estimation.

Returns:

  • positions (ndarray of shape (N, 2)) – The Cartesian coordinates of the sampled k-space points in 2D plane.

  • acs_positions (ndarray of shape (M, 2)) – The Cartesian coordinates of the ACS region k-space points in 2D plane.

Return type:

ndarray[tuple[int, …], dtype[_ScalarType_co]] | tuple[ndarray[tuple[int, …], dtype[_ScalarType_co]], ndarray[tuple[int, …], dtype[_ScalarType_co]]]

Notes

  • This function merely gives the mask positions, not the entire trajectory.