A generative adversarial network to improve integrated mode proton imaging resolution using paired proton–carbon data
This study demonstrates that deep learning can enhance pRad image quality by leveraging paired proton-carbon data. Proton2Carbon can be integrated into existing imaging workflows to improve clinical and research applications of proton radiography. To facilitate further research, the full dataset used to train Proton2Carbon is publicly released and available at https://zenodo.org/records/14945165.
Authors
- Lennart Volz (ORCID: https://orcid.org/0000-0003-0441-4350)
- Christian Graeff (ORCID: https://orcid.org/0000-0002-5296-7649)
- Soon‐Cheol Chung (ORCID: https://orcid.org/0000-0002-4693-5157)
- Ryan Fullarton (ORCID: https://orcid.org/0000-0001-7817-1596)
- Mikaël Simard (ORCID: https://orcid.org/0000-0003-0610-6319)
- Christoph Schuy (ORCID: https://orcid.org/0000-0001-8407-9358)
- Charles‐Antoine Collins‐Fekete (ORCID: https://orcid.org/0000-0003-2464-8795)
- Colin Baker (ORCID: https://orcid.org/0009-0005-7293-3330)
Institutions
- University College London Hospitals NHS Foundation Trust (GB)
- GSI Helmholtz Centre for Heavy Ion Research (DE)
- University College London (GB)
Publication Details
- Journal
- Medical Physics
- Published
- 2025-09-01
- DOI
- https://doi.org/10.1002/mp.18081
- Citations
- 1
- Primary Topic
- Radiation Therapy and Dosimetry
- Type
- article
- Field-Weighted Citation Impact
- 1.52
Funders
- Philipps-Universität Marburg
- UK Research and Innovation
- Cancer Research UK
- National Institute for Health and Care Research