DoseRAD2026 Challenge dataset: AI accelerated photon and proton dose calculation for radiotherapy
2026
Read DoseRAD2026 Challenge dataset: AI accelerated photon and proton dose calculation for radiotherapy (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-12778.txt.
Introduction
Fan Xiao and 15 colleagues present the DoseRAD2026 dataset for radiotherapy dose calculation. Accurate doses spare healthy tissue while hitting tumors. The set pairs CT and MRI scans from 115 patients. Photon and proton dose maps come from Monte Carlo methods, with 40,500 photon beams and 81,000 proton beamlets. Files use compressed MetaImage format with JSON beam settings. Training data are public from April 2026; test data stay hidden until March 2030. The benchmark supports fast dose methods for both therapies.
Catalog record
| YEAR | 2026 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42138 |
| LICENSE | http://creativecommons.org/licenses/by-sa/4.0/ |
| RIGHTS | CC BY SA |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.12778
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
DoseRAD2026 Challenge dataset: AI accelerated photon and proton dose calculation for radiotherapy Abstract Purpose: Accurate dose calculation is a fundamental component of modern radiotherapy, enabling precise tumor irradiation while sparing surrounding healthy tissue. With the increasing adoption of \ac mri-guided radiotherapy and growing interest in real-time adaptive radiotherapy, there is a demand for fast and accurate dose calculation techniques on both \ac ct and \ac mri. To facilitate research in this area, the DoseRAD2026 dataset and accompanying challenge introduce a publicly available benchmark dataset comprising paired CT and MRI data with corresponding beam-level photon and proton \ac mc dose distributions, enabling the development and systematic evaluation of advanced dose calculation and prediction methods. Acquisition and validation methods: The DoseRAD2026 dataset comprises paired CT and MRI data from 115 patients (75 training, 40 testing) treated on a hybrid \ac mri-linear accelerator ( \ac mri-linac) for thoracic or abdominal lesions and is derived from the SynthRAD2025 dataset. Image pre-processing included deformable image registration, air-cavity correction and resampling. Ground-truth photon (6 MV) and proton dose distributions were generated using open-source \ac mc dose calculation algorithms and resulted in a training dataset of 40,500 photon beams and 81,000 proton beamlets. Data format and usage notes: The dataset is organized into photon and proton subsets, each comprising paired \ac ct- \ac mri images, beam-level dose distributions, and JSON files specifying patient-specific beam configurations. Data are provided in compressed MetaImage (.mha) format with standardized naming linking dose files to beam parameters in the JSON files. The dataset is released under a CC BY-NC 4.0 license, with the training set available from April 2026 and the test set withheld until March 2030. Potential applications: The DoseRAD2026 dataset enables benchmarking and development of fast dose calculation methods for radiotherapy applications. Potential use cases include rapid beam-level dose estimation for photon and proton therapy, \ac mri-based dose calculation in \ac mri-only and \ac mri-guided radiotherapy workflows, and support for real-time adaptive radiotherapy strategies. Fan Xiao 1 , Nikolaos Delopoulos 1 , Niklas Wahl 2,3 , Lennart Volz 4 , Lina Bucher 5,6,2,3 , Matteo Maspero 7 , Miguel Palacios 8 , Muheng Li 9 , Samir Schulz 5,10,2,3 , Viktor Rogowski 11 , Ye Zhang 9 , Zoltan Perko 12,13 , Christopher Kurz 1 , George Dedes 1,14 , Guillaume Landry 1,15,16 , Adrian Thummerer 1,17,18,a 1 Department of Radiation Oncology, LMU University Hospital, LMU Munich, Munich, Germany 2 Division of Medical Physics in Radiation Oncology, German Cancer Research Center (DKFZ), Heidelberg, Germany 3 National Center for Radiation Research in Oncology (NCRO), Heidelberg Institute for Radiation Oncology (HIRO), Heidelberg, Germany 4 GSI Helmholtzzentrum für Schwerionenforschung, Biophysics Department, Darmstadt, Germany 5 Department of Physics and Astronomy, Heidelberg University, Heidelberg, Germany 6 Institute of Biomedical Engineering, Karlsruhe Institute of Technology, Baden-Württemberg, Germany 7 Department of Radiotherapy, University Medical Center Utrecht, Utrecht, Netherlands 8 Department of Radiation Oncology, Amsterdam UMC location Vrije Universiteit Amsterdam, Cancer Center Amsterdam, Cancer Treatment and Quality of Life, Amsterdam, Netherlands 9 Center for Proton Therapy, Paul Scherrer Institute (PSI), Villigen, Switzerland 10 Department of Radiation Oncology, University Hospital Heidelberg, Heidelberg, Germany 11 Radiation Physics, Department of Hematology, Oncology, and Radiation Physics, Skåne University Hospital, Lund, Sweden 12 Delft University of Technology, Department of Radiation Science and Technology, Delft, Netherlands 13 Radformation Inc., New York, US 14 Department of Medical Physics, Faculty of Physics, Ludwig-Maximilians-Universität München, Munich, Germany 15 German Cancer Consortium (DKTK), partner site Munich, a partnership between DKFZ and LMU University Hospital, Germany 16 Bavarian Cancer Research Center (BZKF), Munich, Germany 17 Department of Radiation Oncology, Inselspital, Bern University Hospital and University of Bern, Bern, Switzerland 18 Department of Digital Medicine, Faculty of Medicine, University of Bern, Bern, Switzerland. a Author to whom correspondence should be addressed: [email protected] Keywords: Dose calculation, Dose prediction, Deep learning, MRI-only \acresetall I . Introduction Radiation therapy is indicated for approximately half of all cancer patients, spanning curative, adjuvant, and palliative contexts 5 , 20 . Photon therapy, such as \ac vmat 28 , enables highly conformal dose distributions around the target volume through continuous gantry rotation combined with dynamic \ac mlc shaping, while sparing surrounding \acp oar. Proton therapy, such as \ac impt 24 , offers a physically distinct advantage through the Bragg peak 45 , whereby the majority of energy is deposited at a finite tissue depth, affording superior sparing of \acp oar distal to the tumor. In both modalities, the accuracy of dose calculation is foundational to treatment plan quality 2 . As workflows increasingly move toward online (immediately prior to radiation delivery) 21 , 29 and ultimately real-time (continuously during radiation delivery) adaptive 25 , 15 , the computational demand placed on dose engines has grown substantially. \ac mc simulation 35 on high quality \ac ct is broadly accepted as the reference standard for dose calculation, owing to its accurate physics modelling of particle transport. However, general-purpose \ac mc implementations suffer from long computation times, with plan-level runtimes spanning hours to tens of hours on a single \ac cpu 30 . \Ac gpu acceleration has substantially narrowed this gap for dose calculations, with recent \ac gpu-based \ac mc engines reporting plan-level runtimes in the range of a few to tens of seconds 8 , 13 , 23 . In parallel, \ac dl methods have emerged as a complementary strategy, demonstrating sub-second dose prediction and showing early promise at \ac mc-comparable accuracy in both proton 27 , 47 , 32 , 26 and photon 17 , 10 , 31 , 46 therapy. Nevertheless, achieving \ac mc-level accuracy and the computational speed required for online or real-time plan adaptation—ideally in the sub-second or millisecond time-frame—remains an open challenge. \Ac mri offers superior soft-tissue contrast and eliminates ionizing radiation exposure compared to \ac ct 37 , making it valuable for target and \ac oar delineation as well as motion management in online adaptive radiotherapy. The clinical introduction of hybrid \ac mr-linac systems 18 , 16 has enabled high-quality soft-tissue imaging prior to and during photon treatment delivery. Meanwhile, prototype systems integrating proton delivery with in-room \ac mri, as well as \ac mri-only proton workflows 12 , 33 , have also been proposed. However, since \ac mri does not provide electron density or material composition information, dose calculation in both modalities continues to rely on \ac ct. Current clinical practice bridges this gap via \ac sct generation 4 , 40 , but this intermediate step introduces additional accuracy potential sources of degradation (difference between \ac sct and CT), computational latency, and overhead to the dose calculation pipeline, remaining a bottleneck for true real-time plan adaptation. Direct \ac dl-based dose prediction from \ac mri offers a conceptually cleaner alternative, collapsing the \ac sct intermediate into a single end-to-end inference step. Early investigations for both photon 22 , 48 and proton 19 , 42 modalities have demonstrated dosimetric feasibility, yet progress has been constrained by the absence of large-scale paired datasets providing beam-level dose ground truth on both \ac ct and \ac mri. To address this gap, we introduce the DoseRAD2026 dataset, the first publicly available external radiotherapy dose dataset providing paired \ac cts and \ac mris (adapted from the SynthRAD2025 dataset 41 ) alongside beam-level photon (6 MV) and proton \ac mc