pytomography.utils.scatter#

Module Contents#

Functions#

get_smoothed_scatter(scatter, proj_meta[, ...])

Smooths SPECT projection metadata

compute_EW_scatter(projection_lower, projection_upper, ...)

Computes triple energy window estimate from lower and upper scatter projections as well as window widths
pytomography.utils.scatter.get_smoothed_scatter(scatter, proj_meta, sigma_theta=0, sigma_r=0, sigma_z=0, N_sigmas=3, FOV_mask=None)[source]#

Smooths SPECT projection metadata

Parameters:

  • scatter (torch.Tensor) – Input projection data

  • proj_meta (ProjMeta) – Projection metadata

  • sigma_theta (float, optional) – Smoothing in theta (specified in degrees). Defaults to 0.

  • sigma_r (float, optional) – Smoothing in r (specified in cm). Defaults to 0.

  • sigma_z (float, optional) – Smoothing in z (specified in cm). Defaults to 0.

  • N_sigmas (int, optional) – Number of sigmas to include in the smoothing kernel. Defaults to 3.

  • FOV_mask (torch.Tensor, optional) – Field of view mask. Defaults to None.

Returns:

Smoothed projections

Return type:

torch.Tensor

pytomography.utils.scatter.compute_EW_scatter(projection_lower, projection_upper, width_lower, width_upper, width_peak, weighting_lower=0.5, weighting_upper=0.5, proj_meta=None, sigma_theta=0, sigma_r=0, sigma_z=0, N_sigmas=3, return_scatter_variance_estimate=False, fov_mask=None)[source]#

Computes triple energy window estimate from lower and upper scatter projections as well as window widths

Parameters:

  • projection_lower (torch.Tensor) – Projection data corresponding to lower energy window

  • projection_upper (torch.Tensor) – Projection data corresponding to upper energy window

  • width_lower (float) – Width of lower energy window

  • width_upper (float) – Width of upper energy window

  • width_peak (float) – Width of peak energy window

  • return_scatter_variance_estimate (bool, optional) – Return scatter variance estimate. Defaults to False.

  • fov_mask (torch.tensor, optional) – Field of view mask; may be required when blurring scatter to avoid blurring accross edges. Defaults to None.

  • weighting_lower (float) –

  • weighting_upper (float) –

  • sigma_theta (float) –

  • sigma_r (float) –

  • sigma_z (float) –

  • N_sigmas (int) –

Returns:

Scatter estimate (and scatter variance estimate if specified)

Return type:

torch.Tensor | Sequence[torch.Tensor]