deepmr.bloch.FSEStep

Contents

deepmr.bloch.FSEStep#

deepmr.bloch.FSEStep(states, ESP, T1, T2, weight=None, k=None, chemshift=None, D=None, v=None, grad_props=None)[source]#

(Unbalanced) SSFP propagator.

Consists of rfpulse (omitted) - free precession - spoiling gradient.

Parameters:
  • states (dict) – EPG states matrix.

  • ESP (float) – Echo Spacing in [ms].

  • T1 (torch.Tensor) – T1 relaxation time of shape (..., npools) in ``[ms].

  • T2 (torch.Tensor) – T2 relaxation time of shape (..., npools) in ``[ms].

  • weight (torch.Tensor) – Pool relative fraction.

  • k (torch.Tensor) – Chemical exchange matrix (...., npools, npools) in [Hz].

  • chemshift (torch.Tensor) – Chemical shift of each pool of shape ``(npools,).

  • D (torch.Tensor) – Apparent diffusion coefficient of shape (...,) in [um**2/ms]. Assume same coefficient for each pool.

  • v (torch.Tensor) – Spin velocity of shape (...,) in [cm/s]. Assume same coefficient for each pool.

  • grad_props (dict) – Extra parameters.

Returns:

  • TEop (deepmr.bloch.Operator) – Propagator until TE. If TE=0, this is the Identity.

  • TETRop (deepmr.bloch.Operator) – Propagator until next TR.