Dose calculation via Transformers
This repository contains the code to calculate the dose deposited by a mono-energetic beam of protons, for arbitrary patient geometries and beam energies.
- GNU General Public License 3.0
- Copyright: Oscar Pastor-Serrano, TU Delft
The 'main' branch contains the code for proton beamlets, while the 'dev_photons' branch contains the code for predictin full photon beams.
Credits
If you like this repository, please click on Star!
If you use the code for your research, please consider citing:
Pastor-Serrano, O., & Perko, Z. (2021). Learning the Physics of Particle Transport via Transformers. https://doi.org/10.1609/aaai.v36i11.21466
Pastor-Serrano, O., & Perko, Z. (2022). Millisecond speed deep learning based proton dose calculation with Monte Carlo accuracy. https://doi.org/10.1088/1361-6560/ac692e
Pastor-Serrano, O., Dong, P., Huang, C., Xing, L., & Perko, Z. (2023). Sub-second photon dose prediction via transformer neural networks. https://doi.org/10.1002/mp.16231
This project is supported by the following institutions:
- KWF Kanker Bestrijding
- Department of Radiation Science and Technology (TU Delft)
Requirements
- Tensorflow 2.5 or higher
- pymedphys
- tensorflow-addons