PET tutorial modification, clinical example added, object space / SPE…

…CT projection space batch removal

.readthedocs.yaml

@@ -3,7 +3,7 @@ version: 2
 build:
   os: "ubuntu-20.04"
   tools:
-    python: "3.9"
+    python: "3.11"
 
 sphinx:
   configuration: docs/source/conf.py

docs/source/autoapi/index.rst

@@ -0,0 +1,11 @@
+API Reference
+=============
+
+This page contains auto-generated API reference documentation [#f1]_.
+
+.. toctree::
+   :titlesonly:
+
+   /autoapi/pytomography/index
+
+.. [#f1] Created with `sphinx-autoapi <https://github.com/readthedocs/sphinx-autoapi>`_

docs/source/autoapi/pytomography/algorithms/dip_recon/index.rst

@@ -0,0 +1,56 @@
+:py:mod:`pytomography.algorithms.dip_recon`
+===========================================
+
+.. py:module:: pytomography.algorithms.dip_recon
+
+
+Module Contents
+---------------
+
+Classes
+~~~~~~~
+
+.. autoapisummary::
+
+   pytomography.algorithms.dip_recon.DIPRecon
+
+
+
+
+.. py:class:: DIPRecon(likelihood, prior_network, rho = 0.003)
+
+   Implementation of the Deep Image Prior reconstruction technique (see https://ieeexplore.ieee.org/document/8581448). This reconstruction technique requires an instance of a user-defined ``prior_network`` that implements two functions: (i) a ``fit`` method that takes in an ``object`` (:math:`x`) which the network ``f(z;\theta)`` is subsequently fit to, and (ii) a ``predict`` function that returns the current network prediction :math:`f(z;\theta)`. For more details, see the Deep Image Prior tutorial.
+
+   :param likelihood: Initialized likelihood function for the imaging system considered
+   :type likelihood: Likelihood
+   :param prior_network: User defined prior network that implements the neural network :math:`f(z;\theta)` that predicts an object given a prior image :math:`z`. This network also implements a ``fit`` method that takes in an object and fits the network to the object (for a specified number of iterations: SubIt2 in the paper).
+   :type prior_network: nn.Module
+   :param rho: Value of :math:`\rho` used in the optimization procedure. Larger values of :math:`rho` give larger weight to the neural network, while smaller values of :math:`rho` give larger weight to the EM updates. Defaults to 1.
+   :type rho: float, optional
+
+   .. py:method:: _compute_callback(n_iter, n_subset)
+
+      Method for computing callbacks after each reconstruction iteration
+
+      :param n_iter: Number of iterations
+      :type n_iter: int
+      :param n_subset: Number of subsets
+      :type n_subset: int
+
+
+   .. py:method:: __call__(n_iters, subit1, n_subsets_osem=1, callback=None)
+
+      Implementation of Algorithm 1 in https://ieeexplore.ieee.org/document/8581448. This implementation gives the additional option to use ordered subsets. The quantity SubIt2 specified in the paper is controlled by the user-defined ``prior_network`` class.
+
+      :param n_iters: Number of iterations (MaxIt in paper)
+      :type n_iters: int
+      :param subit1: Number of OSEM iterations before retraining neural network (SubIt1 in paper)
+      :type subit1: int
+      :param n_subsets_osem: Number of subsets to use in OSEM reconstruction. Defaults to 1.
+      :type n_subsets_osem: int, optional
+
+      :returns: Reconstructed image
+      :rtype: torch.Tensor
+
+
+

docs/source/autoapi/pytomography/algorithms/fbp/index.rst

@@ -0,0 +1,44 @@
+:py:mod:`pytomography.algorithms.fbp`
+=====================================
+
+.. py:module:: pytomography.algorithms.fbp
+
+.. autoapi-nested-parse::
+
+   This module contains classes that implement filtered back projection reconstruction algorithms.
+
+
+
+Module Contents
+---------------
+
+Classes
+~~~~~~~
+
+.. autoapisummary::
+
+   pytomography.algorithms.fbp.FilteredBackProjection
+
+
+
+
+.. py:class:: FilteredBackProjection(projections, system_matrix, filter=None)
+
+   Implementation of filtered back projection reconstruction :math:`\hat{f} = \frac{\pi}{N_{\text{proj}}} \mathcal{R}^{-1}\mathcal{F}^{-1}\Pi\mathcal{F} g` where :math:`N_{\text{proj}}` is the number of projections, :math:`\mathcal{R}` is the 3D radon transform, :math:`\mathcal{F}` is the 2D Fourier transform (applied to each projection seperately), and :math:`\Pi` is the filter applied in Fourier space, which is by default the ramp filter.
+
+   :param projections: projection data :math:`g` to be reconstructed
+   :type projections: torch.Tensor
+   :param system_matrix: system matrix for the imaging system. In FBP, phenomena such as attenuation and PSF should not be implemented in the system matrix
+   :type system_matrix: SystemMatrix
+   :param filter: Additional Fourier space filter (applied after Ramp Filter) used during reconstruction.
+   :type filter: Callable, optional
+
+   .. py:method:: __call__()
+
+      Applies reconstruction
+
+      :returns: Reconstructed object prediction
+      :rtype: torch.tensor
+
+
+