portpy 0.0.5

First open-source radiation treatment planning system in Python

What is PortPy?

Note: The package is at its early stages of development (version 0.0.1) and we are now collecting feedback from researchers to further refine the data structure and the main functionality. We are expecting to have a stable version 1.xx around July 2023. We would love to hear your feedback.

PortPy (Planning and Optimization for Radiation Therapy) is a community effort to develop the first opensource python library to facilitate the development and clinical translation of radiotherapy cancer treatment planning algorithms. PortPy includes:

1. Research-ready data and code for benchmarking, reproducibility, and community-driven development.
2. Interface to an open-source optimization package CVXPy for easy/quick prototyping and out-of-the-box access to commercial/opensource optimization engines (e.g., Mosek, Gorubi, CPLEX).
3. Visualization modules to visualize relevant plan information (e.g, dose volume histograms, dose distribution, fluence map).
4. Evaluation modules to quantify plan quality with respect to established clinical metrics (e.g., RTOG metrics, dose conformality, tumor control probability, normal tissue control probability).

Data

Data needed for optimization and algorithm development (e.g., a set of beams/beamlets/voxels, dose contribution of each beamlet to each voxel) are provided for a set of pre-specified machine parameters (e.g., beam/collimator/couch angles). We will initially provide these for a set of publicly available datasets from TCIA. We initially start with two lung patients to get feedback and then expand to about 100 patients in version 1.xx (about end of July). We hope to further expand our dataset in the future. The data needed for optimization is extracted from Eclipse^TM, a FDA-approved commerical treatment planning system (Varian Medical Systems), using its API.

You can download the sample patient data here.

Create a directory named './data' in the current project directory and copy the downloaded file to it, e.g ./data/Lung_Phantom_Patient_1

Quick Start

Please see below for understanding the basic functionalities of PortPy. For advance usage of PortPy, we recommend navigating through examples folder.

1. To understand the most important features of PortPy, we highly recommend going through notebook ex_1_introduction.ipynb
2. One of the major computational issues while optimizing the plan arise due to large size of influence matrix. We suggest you to follow ex_2_down_sampling.py and ex_4_inf_matrix_sparsification.py to understand how PortPy can assist in resolving it.
3. You can check out ex_3_structure_operations.py to know how to perform different structure operations (e.g., boolean, margin).
4. For algorithm benchmarking, the global optimal solution is provided for non-convex optimization problems resulting from beam angle optimization ex_6_boo_benchmark.ipynb, incorporating DVH constraints ex_5_dvh_benchmark.py, and VMAT optimization ex_8_VMAT.py using the mixed-integer programming on down-sampled data.
5. In addition to basic visualization capabilities, PortPy provide advanced visualization by integration with 3D Slicer. Please check out notebook ex_7_Slicer.ipynb

Installing PortPy

1. Installing using pip

pip install portpy

2. Installing using conda

conda install -c conda-forge portpy

3. Installing from source

• Clone this repository:

  git clone https://github.com/PortPy-Project/PortPy.git
  cd portpy
  

• You need to install the dependencies in either a python virtual environment or anaconda environment. Instructions for setting up in python virtual environment are as follows:

  Install all the dependencies present in requirements.txt:

  python3 -m venv venv
  source venv/bin/activate
  (venv) pip install -r requirements.txt
  

Team

PortPy is a community project initiated at Memorial Sloan Kettering Cancer Center. It is currently developed and maintained by:

Name	Expertise	Institution
Masoud Zarepisheh	Treatment Planning and Optimization	MSK
Saad Nadeem	Computer Vision and AI in Medical Imaging	MSK
Gourav Jhanwar	Algorithm Design and Development	MSK
Mojtaba Tefagh	Mathematical Modeling and Reinforcement Learning	SUT
Vicki Taasti	Physics and Planning of Proton Therapy	MAASTRO
Sadegh Alam	Adaptive Treatment Planning and Imaging	Cornell
Seppo Tuomaala	Eclispe API Scripting	VARIAN

License

PortPy code is distributed under Apache 2.0 with Commons Clause license, and is available for non-commercial academic purposes.