PCaNet

SPDX-FileCopyrightText: Copyright (C) 2021-2023 Asmail Muftah [email protected], PhD student at Cardiff University
SPDX-FileCopyrightText: Copyright (C) 2020-2024 Frank C Langbein [email protected], Cardiff University
SPDX-FileCopyrightText: Copyright (C) 2020-2022 S Shermer [email protected], Swansea University
SPDX-License-Identifier: AGPL-3.0-or-later

Code for classification and segmentation of prostate cancer.

The main functions are in the pcanet module. We provide jupyter notebook and command line interfaces, as needed/used. The results are in separate repositories:

• https://qyber.black/ca/results-pcanet-models-classification Some repositories are submodules. Note that they may be quite large.

Also note that the data repositories may not be public due to distribution limitations.

Install Instructions (Linux)

These instructions are for Linux. Note that the code has only been tested on Linux and may not work on another platform.

1. Clone the repository:

   git clone [email protected]:ca/pca/code-pcanet.git pcanet
   

   Check the clone url, as it may be different if you use a different repository, e.g. from a mirror or alternative versions for development, etc.
2. Navigate to the directory:

   cd pcanet
   

   Make sure to select a branch or tag with git checkout BRANCH_OR_TAG for a specific version instead of the main branch.
3. Update submodules:

   git submodule update --init --recursive
   

   Note, some of the sub-modules are quite large and may not be accessible (if we cannot distribute the data; sometimes you may be able to get the data from the original source and use the tools from code-qdicom-utilities to convert the original files to a suitable data format). It may also be best to set GIT_LFS_SKIP_SMUDGE=1 to avoid the large data files are downloaded via LFS and then pull only specific LFS files with git lfs pull -I PATH that you need.
4. Install the requirements:

   pip3 install -r requirements.txt
   

   Note that the requirements may need additional libraries, etc. to be installed on you system that pip does not add automatically. Also note, that some packages (in particular mahotas), may not be available at the requested version if you are using a more recent version of python (python3.10) - these packages can still be installed from source; obtain the sources and checkout the version requested and run python3 setup.py install --prefix=~/.local or similar.

Command Line Interface

• pca.py is the command line interface to the code for some functionalities. See --help option for further information.
  • augment sub-command creates augmented datasets for classification and segmentation. Classification includes extracting patches cropped to a rectangular region. Default augmentation options are set in pcanet/augment.py for the two modes. The original sample is always included. If the augmentation factor is 1, then only the original sample is included.
  • classify trains various classifiers on the datasets. If the model specified already exists it only evaluates it on the dataset given.
  • predict runs a given classifier on a dataset to predict results.
  • view shows augmentation dataset content.

Jupyter Notebooks

None directly included in the package, but see submodules that may have some to analyse the results, etc.