Extending of the pytrip software package for biologically optimized proton radiotherapy

Abstract

Recent years brought several developments in treatment planning with variable radiobiological effectiveness for proton radiotherapy. Commercial clinical and research treatment planning systems are, in addition to their costs and availability, hard to adapt for emerging models and new approaches in proton therapy planning. We aim at providing suitable, open access and extensible software toolbox for research in treatment planning, based on the TRiP98 program and freely available libraries. The research platform is composed of two projects: a core library called pytrip and a graphical user interface (GUI) called pytripgui. Both projects are implemented mainly in the Python programming language. The core library is capable of handling DICOM and VOXELPLAN data format by custom readers based on the NumPy library. Pytrip provides a powerful toolbox for research in treatment planning. In our system, the treatment plan is calculated by the TRiP98 program which can be run locally or remotely. Several additions not included in TRiP98 are provided by the pytrip toolbox, like empirical models of variable RBE in proton therapy, calculations of DVH, DICOM support. Graphical user interface supplements TRiP98 by a plotting tool, capable of handling CT scans and relevant quantities like dose, LET and RBE. GUI source code was upgraded to use a new widget toolkit - PyQt5 framework, which enables maintenance and simplifies further development of the project. Pytrip provides a powerful toolbox for research in treatment planning. It enables the development and testing of variable RBE models for proton therapy. Graphical user interface aids users in the visualization and calculation of treatment plans for research purposes.