PETSIRD draft data model and examples

The purpose of this repo is to have a working example of a data model for PET imaging (list mode data). This is not complete.

Background

The Emission Tomography Standardization Initiative (ETSI) is working towards establishing a standard for PET Raw Data, called PETSIRD ("PET ETSI Raw Data").

The specification uses the yardl tool to define the model. yardl can be used to read the specification (in the model directory) and generate an SDK for both C++ and Python to read/write PETSIRD data.

To get started as a Python user:

If you don't want to modify the yardl model, just pip install petsird. More detail is in the python/README.md.

To get started as a C++ user:

You can use conda install conda-forge::libpetsird.

To get started quickly as a developer:

1. Open this repo in GitHub Codespaces or in a VS Code devcontainer. This codespace/container will contain all necessary tools, including yardl itself, as well as the current repository. Alternatively, you can also create a conda environment, run the install_yardl.sh script, and add $PWD/yardl to your PATH.

2. Browse the model directory.

3. Use yardl to generate C++ and Python code for the model and compile/install:

   cd /wherever/PETSIRD
   just build

4. Have a look at (and try!) the examples in the cpp and/or python directories.

PETSIRD maintainer notes

• Use pre-commit before you push to GitHub.

• If it is necessary to update yardl, edit the install_yardl.sh script.

• When creating a new release, you need to

  1. update the version in cpp/CMakeLists.txt (sorry)
  2. merge to main
  3. run

  VER=vx.x.x
  git tag -a ${VER} -m ${VER}
  git push origin ${VER}

  4. edit the draft release automatically created on GitHub and publish it