What we do
About our project
The Radiotherapy Optimisation Test Set (TROTS) is an extensive set of problems originating from radiation therapy treatment planning [1,2].
When a patient is diagnosed with cancer and selected for treatment with radiation therapy, a so-called treatment plan has to be generated. This is based on a 3D or 4D Computer Tomography (CT) scan of the patient. The treatment plan describes the personalised settings of the applied treatment unit, and contains a predicted patient dose distribution for these settings, projected on the planning CT-scan. The delivered dose distribution determines both the probability for cure and also the probability of damage to healthy tissues. The aim is to deliver sufficient dose to the tumour for curation, while keeping the dose to healthy organs as low as possible to minimise the probability of developing radiation-induced treatment related complications. More background information can be found on the pages of Radiation oncology and the website of Sebastiaan Breedveld.
Computing a treatment plan is, at its worst, a large-scale non-convex non-linear combinatorial multi-crtiterial optimisation problem. In  several techniques are discussed to transform this in a series of single-criteria optimisation models. As short planning are relevant, the problems should be solved as fast as possible. And to acceptable accuracy: suboptimal results from the solver may lead to suboptimal treatment plans, meaning that the patient will not receive the best treatment technically possible.
TROTS provides a variety of cases for different patients and treatment sites. This dataset can be used to test and evaluate the performance of numerical solvers  or to investigate multi-criteria optimisation and decision-making . Image sets and scripts to visualise the data are also provided, as these are important tools for proper interpretation of the results, especially from a decision-making point of view . The format and details on the dataset are described in .
Visualisation of the data in Matlab - easy dose-volume histograms and dose distribution
[IMAGE CTAnon.jpg : All data is properly anonymised, including CT data and facial profile.]
Data locations and descriptions
Alternatively, the data can be accessed directly by FTP:
If there are any issues with downloading, please contact: firstname.lastname@example.org
In these problems, Prostate_CK contains 30 patients using a fixed 25-beam setup and is described in [4,5]. Prostate_VMAT containts 30 other prostate patients using a different protocol for prostate cancer, and is a 23-beam setup to mimic a VMAT dose distribution [6,7]. Head-and-Neck and Head-and-Neck_Alt both contain the same 15 patients for a 23-beam VMAT approximation where the _Alt contains higher accuracy dose model, resulting in denser (thus heavier) problems [8,9]. Protons contains 20 patients, to be treated with 3-beam proton therapy [10,11], and are formulated as linear problems. The proton background also results in a different type of data matrices, compared to the other photon-based techniques. The Liver set containts 10 patients including non-convex constraints, using a 15 beam non-coplanar optimised beam setup . The Prostate_BT set contains 25 high dose rate brachytherapy prostate cases, where the prescription is alsmost exclusively based on the non-convex dose-volume criterion .
The description of the data format (Matlab v7.3 HDF5 files, see ) and numerical results (obtained by the Erasmus-iCycle solver, see ) can be downloaded here:
1. Breedveld S., van den Berg B. & Heijmen B. (2017) An interior-point implementation developed and tuned for radiation therapy treatment planning Comput. Optim. Appl. 68 209-242 (DOI)
2. Breedveld S., Craft D., van Haveren R. & Heijmen B. (2019) Multi-criteria optimisation and decision-making in radiotherapy Eur. J. Oper. Res. 277 1-19 (DOI)
3. Breedveld S. & Heijmen B. (2017) Data for TROTS - The Radiotherapy Optimisation Test Set Data in Brief 12 143-149 (TROTS location, DOI)
4. Rossi L., Breedveld S., Aluwini S. & Heijmen B. (2015) Non-coplanar beam angle class solutions to replace time-consuming patient-specific beam angle optimization in robotic prostate SBRT Int. J. Radiat. Oncol. Biol. Phys. 92 762-770 (DOI)
5. Rossi L., Breedveld S., Heijmen B., Voet P., Lanconelli N. & Aluwini S. (2012) On the beam direction search space in computerized non-coplanar beam angle optimization for IMRT -- prostate SBRT Phys. Med. Biol. 57 5441-5458 (DOI)
6. Voet P., Dirkx M., Breedveld S., Al-Mamgani A., Incrocci L. & Heijmen B. (2014) Fully automated VMAT plan generation for prostate cancer patients Int. J. Radiat. Oncol. Biol. Phys. 88 1175-1179 (DOI)
7. Van Haveren R., Breedveld S., Keijzer M., Voet P., Heijmen B. & Ogryczak W. (2017) Lexicographic Extension of the Reference Point Method Applied in Radiation Therapy Treatment Planning Eur. J. Oper. Res. 263 247–257 (DOI)
8. Voet P., Breedveld S., Dirkx M., Levendag P. & Heijmen B. (2012) Integrated multicriterial optimization of beam angles and intensity profiles for coplanar and noncoplanar head and neck IMRT and implications for VMAT Med. Phys. 39 4858-4865 (DOI)
9. Van Haveren R., Ogryczak W., Verduijn G., Keijzer M., Heijmen B. & Breedveld S. (2017) Fast and fuzzy multi-objective radiotherapy treatment plan generation for head-and-neck cancer patients with the lexicographic reference point method (LRPM) Phys. Med. Biol. 62 4318 (DOI)
10. Van de Water S., Kraan A., Breedveld S., Schillemans W., Teguh D., Kooy H., Madden T., Heijmen B. & Hoogeman M. (2013) Improved efficiency of multi-criteria IMPT treatment planning using iterative resampling of randomly placed pencil beams Phys. Med. Biol. 58 6969 (DOI)
11. Van de Water S., Kooy H., Heijmen B. & Hoogeman M. (2015) Shortening delivery times of intensity modulated proton therapy by reducing proton energy layers during treatment plan optimization IJROBP. 92 460-468 (DOI)
12. Breedveld S., Storchi P. Voet P. & Heijmen B. (2012) iCycle: Integrated, multicriterial beam angle, and profile optimization for generation of coplanar and noncoplanar IMRT plans Med. Phys. 39 951-963 (DOI)
13. Breedveld S., Bennan A., Aluwini S., Schaart D., Kolkman-Deurloo I-K. & Heijmen B. (2019) Fast automated multi-criteria planning for HDR brachytherapy explored for prostate cancer submitted