Source code for pytomography.metadata.SPECT.spect_metadata

from __future__ import annotations
from typing import Sequence
import pytomography
from pytomography.utils import compute_pad_size
import torch
from..metadata import ObjectMeta, ImageMeta

[docs]class SPECTObjectMeta(ObjectMeta): """Metadata for object space in SPECT imaging Args: dr (list[float]): List of 3 elements specifying voxel dimensions in cm. shape (list[int]): List of 3 elements [Lx, Ly, Lz] specifying the length of each dimension. """ def __init__(self, dr: list[float], shape: list[int]) -> None: self.dr = dr self.dx = dr[0] self.dy = dr[1] self.dz = dr[2] self.shape = shape self.compute_padded_shape() self.recon_method = None self.units = 'counts'
[docs] def compute_padded_shape(self) -> list: """Computes the padded shape of an object required when rotating the object (to avoid anything getting cut off). """ self.pad_size = compute_pad_size(self.shape[0]) x_padded = self.shape[0] + 2*self.pad_size y_padded = self.shape[1] + 2*self.pad_size z_padded = self.shape[2] self.padded_shape = (int(x_padded), int(y_padded), int(z_padded))
[docs]class SPECTImageMeta(ImageMeta): """Metadata for image space in SPECT imaging Args: projection_shape (Sequence): 2D shape of the projections angles (Sequence): Specifies the detector angles for all projections in image space radii (Sequence, optional): Specifies the radial distance of the detector corresponding to each angle in `angles`; only required in certain cases (i.e. PSF correction). Defaults to None. """ def __init__( self, projection_shape: Sequence, angles: Sequence, radii=None ) -> None: self.angles = torch.tensor(angles).to(pytomography.device).to(pytomography.dtype) self.radii = radii self.num_projections = len(angles) self.shape = (self.num_projections, projection_shape[0], projection_shape[1]) self.compute_padded_shape()
[docs] def compute_padded_shape(self) -> list: """Computes the padded shape of an object required when rotating the object (to avoid anything getting cut off). """ self.pad_size = compute_pad_size(self.shape[1]) theta_padded = self.shape[0] r_padded = self.shape[1] + 2*self.pad_size z_padded = self.shape[2] self.padded_shape = (int(theta_padded), int(r_padded), int(z_padded))
[docs]class SPECTPSFMeta(): r"""Metadata for PSF correction. PSF blurring is implemented using Gaussian blurring with :math:`\sigma(r) = f(r,p)` where :math:`r` is the distance from the detector, :math:`\sigma` is the width of the Gaussian blurring at that location, and :math:`f(r,p)` is the ``sigma_fit`` function which takes in additional parameters :math:`p` called ``sigma_fit_params``. (By default, ``sigma_fit`` is a linear curve). As such, :math:`\frac{1}{\sigma\sqrt{2\pi}}e^{-r^2/(2\sigma(r)^2)}` is the point spread function. Blurring is implemented using convolutions with a specified kernel size. Args: sigma_fit_params (float): Parameters to the sigma fit function sigma_fit (function): Function used to model blurring as a function of radial distance. Defaults to a 2 parameter linear model. kernel_dimensions (str): If '1D', blurring is done seperately in each axial plane (so only a 1 dimensional convolution is used). If '2D', blurring is mixed between axial planes (so a 2D convolution is used). Defaults to '2D'. min_sigmas (float, optional): This is the number of sigmas to consider in PSF correction. PSF are modelled by Gaussian functions whose extension is infinite, so we need to crop the Gaussian when computing this operation numerically. Note that the blurring width is depth dependent, but the kernel size used for PSF blurring is constant. As such, this parameter is used to fix the kernel size such that all locations have at least ``min_sigmas`` of a kernel size. """ def __init__( self, sigma_fit_params: Sequence[float, float], sigma_fit : function = lambda r, a, b: a*r+b, kernel_dimensions: str = '2D', min_sigmas: float = 3 ) -> None: self.sigma_fit_params = sigma_fit_params self.sigma_fit = sigma_fit self.kernel_dimensions = kernel_dimensions self.min_sigmas = min_sigmas