bartorch.remove_readout_oversampling

bartorch.remove_readout_oversampling()

Crop the readout’s field of view to target_size samples, in the image domain.

The readout is transformed to the image domain with the centred unitary transform (bartorch.fft()), cropped symmetrically about index n // 2 and transformed back: the result is k-space over the same extent, sampled n / target_size times more coarsely.

Parameters:

• kspace (torch.Tensor) – Centred k-space with the readout along axis.

• target_size (int) – Number of readout samples to keep, the reconstructed matrix size.

• axis (int, default=-1) – Readout axis.

Returns:

Complex64 k-space with target_size samples along axis, or kspace itself when it already has target_size.

Return type:

torch.Tensor

Raises:

ValueError – If target_size is not in [1, n].

Examples using remove_readout_oversampling

Readout oversampling and apodization

Readout oversampling and apodization