bartorch.tools.ecalib

Contents

bartorch.tools.ecalib#

bartorch.tools.ecalib(kspace, *, maps=None, calib_size=None, threshold=None, crop=None, kernel_size=None, softsense=False, intensity_correction=False, return_eigenvalues=False, **extra)#

Coil sensitivities by ESPIRiT.

Parameters:

  • kspace (torch.Tensor) – Fully sampled calibration data, or k-space with a sampled centre.

  • maps (int, optional) – How many sets of sensitivities to produce (-m).

  • calib_size (int or tuple of int, optional) – The calibration region’s size (-r), the same on every axis or one per axis.

  • threshold (float, optional) – The singular-value threshold for the calibration matrix (-t).

  • crop (float, optional) – The eigenvalue below which a sensitivity is set to zero (-c).

  • kernel_size (int, optional) – The calibration kernel’s size (-k).

  • softsense (bool) – Return the maps without the eigenvalue crop, for soft-SENSE (-S).

  • intensity_correction (bool) – Correct for intensity rather than normalising (-I).

  • return_eigenvalues (bool) – Also return the eigenvalue map, which BART writes as a second array only when asked.

  • **extra – Further BART ecalib options, by name. e, the axis the second step is split along, takes an axis of kspace.

Returns:

The sensitivities, and the eigenvalues when asked for.

Return type:

torch.Tensor or tuple of torch.Tensor

Examples

>>> maps = ecalib(kspace, maps=1, crop=0.8)
Contents