Credits
Every scan on this site comes from an open dataset. Each one is credited here with its license.
Code
Ganform code is Apache-2.0. Surface distance definitions follow DeepMind surface-distance 0.1, also Apache-2.0.
Fonts
Geist and Geist Mono are SIL Open Font License 1.1. The license texts ship with the font files. No font service is contacted.
LCTSC
Yang, J., Veeraraghavan, H., Armato, S. G., et al. Data from Lung CT Segmentation Challenge (LCTSC) (Version 3). The Cancer Imaging Archive. https://doi.org/10.7937/K9/TCIA.2017.3R3FVZ08. License: CC BY 3.0.
Clark, K., Vendt, B., Smith, K., Freymann, J., Kirby, J., Koppel, P., Moore, S., Phillips, S., Maffitt, D., Pringle, M., Tarbox, L., and Prior, F. (2013). The Cancer Imaging Archive (TCIA): Maintaining and Operating a Public Information Repository. Journal of Digital Imaging, 26(6), 1045-1057. https://doi.org/10.1007/s10278-013-9622-7
Packages derived from LCTSC keep that citation, CC BY 3.0, and the TCIA data usage policy: https://wiki.cancerimagingarchive.net/x/c4hF. Downstream users must keep the same three. The home page outlines are organ contours from LCTSC-Train-S1-001 on one slice. The app does not offer skin surfaces or 3D volume rendering.
Prostate Anatomical Edge Cases
Thompson, R. F., Kanwar, A., Merz, B., Cohen, E., Fisher, H., Rana, S., Claunch, C., & Hung, A. (2023). Stress-Testing Pelvic Autosegmentation Algorithms Using Anatomical Edge Cases (Prostate Anatomical Edge Cases) (Version 1) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/QSTF-ST65. License: CC BY 4.0.
Kanwar, A., Merz, B., Claunch, C., Rana, S., Hung, A., & Thompson, R. F. (2023). Stress-testing pelvic autosegmentation algorithms using anatomical edge cases. Physics and Imaging in Radiation Oncology, 25, 100413. https://doi.org/10.1016/j.phro.2023.100413
Pelvic cases are cropped from these simulation CT scans and use the bladder, prostate, rectum and femoral head contours that came with them, which the dataset says were used for treatment planning. Packages derived from them keep that citation, CC BY 4.0, the Clark et al. citation above and the TCIA data usage policy: https://wiki.cancerimagingarchive.net/x/c4hF.
LIDC-annot-NLST501
Renoust, B., Baudot, P., Voyton, C. M., Chikbouni, D., Foriel, T., Haddou, Y., Geremia, E., De Bie, G., LE, V. K., Bobin, V., Francis, D., Siot, P.-H., Bodard, S., Bourdes, V., & Huet, B. (2026). LIDC-IDRI-like Multi-Reader annotations of lung nodules on NLST CT scans (LIDC-annot-NLST501) [Dataset]. The Cancer Imaging Archive. https://doi.org/10.7937/Y2DX-E334. License: CC BY 4.0.
National Lung Screening Trial Research Team. (2013). Data from the National Lung Screening Trial (NLST) (Version 3) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/TCIA.HMQ8-J677. License: CC BY 4.0.
Lung nodule cases are cropped around one lesion that at least three radiologists outlined independently. Each reader's outline is kept, and the reference is a majority vote of those readers that Ganform counted itself; it is not a consensus the dataset released. Packages derived from them keep both citations, CC BY 4.0, the Clark et al. citation above and the TCIA data usage policy: https://wiki.cancerimagingarchive.net/x/c4hF.
TotalSegmentator
TotalSegmentator is a CT segmentation tool by Jakob Wasserthal and coauthors. The code and the open task total are Apache-2.0. The TotalSegmentator CT dataset labels are CC BY 4.0.
Wasserthal, J., Breit, H.-C., Meyer, M.T., Pradella, M., Hinck, D., Sauter, A.W., Heye, T., Boll, D., Cyriac, J., Yang, S., Bach, M., and Segeroth, M. (2023). TotalSegmentator: Robust Segmentation of 104 Anatomic Structures in CT Images. Radiology: Artificial Intelligence. https://doi.org/10.1148/ryai.230024
Dataset: Wasserthal, J. (2023). Small subset of TotalSegmentator Dataset (Version 2.0.1) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.10047263. License: CC BY 4.0. Beam Puzzle uses single axial CT slices of the chest, pelvis and abdomen from this subset with its heart, lung, esophagus, spinal cord, bladder, prostate, bowel, femur, liver, kidney, stomach and pancreas labels, which are dataset labels and may be model assisted. Contouring cases, Anatomy Sprint and Level Finder use thorax scans cropped from this subset with its trachea, great vessel, sternum, lung lobe, rib, vertebra, clavicle, scapula, humerus, costal cartilage and thyroid labels. The crop stops below the jaw and the results say Match to the dataset label.
TotalSegmentator is built on nnU-Net, also Apache-2.0: https://github.com/MIC-DKFZ/nnUNet
RIDER-LungCT-Seg, 4D-Lung and Pediatric-CT-SEG
Wee, L., Aerts, H., Kalendralis, P., & Dekker, A. (2020). RIDER Lung CT Segmentation Labels (RIDER-LungCT-Seg) (Version 2) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/tcia.2020.jit9grk8. License: CC BY 3.0. The images are RIDER Lung CT: Zhao, B., Schwartz, L. H., & Kris, M. G. (2015). Data From RIDER Lung CT (Version 2) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/k9/tcia.2015.u1x8a5nr. Each case shows a lung tumor on one of two scans of the same patient. The dataset's manual contour is the reference, and its automatic contour is a review item and the "you vs the dataset's auto-contour" line.
Hugo, G. D., Weiss, E., Sleeman, W. C., Balik, S., Keall, P. J., Lu, J., & Williamson, J. F. (2016). Data from 4D Lung Imaging of NSCLC Patients (Version 2) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/K9/TCIA.2016.ELN8YGLE. License: CC BY 3.0. See also Hugo et al. (2017), Medical Physics 44(2), 762-771, https://doi.org/10.1002/mp.12059. One radiation oncologist contoured every breathing phase; Motion Envelope plays those phases as a loop.
Jordan, P., Adamson, P. M., Bhattbhatt, V., Beriwal, S., Shen, S., Radermecker, O., Bose, S., Strain, L. S., Offe, M., Fraley, D., Principi, S., Ye, D. H., Wang, A. S., Van Heteren, J., Vo, N.-J., & Schmidt, T. G. (2021). Pediatric Chest/Abdominal/Pelvic CT Exams with Expert Organ Contours (Pediatric-CT-SEG) (Version 2) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/TCIA.X0H0-1706. License: CC BY 4.0. See also Jordan et al. (2022), Medical Physics 49(5), 3523-3528, https://doi.org/10.1002/mp.15485. Only the chest structures are used here, and these cases stay out of the Daily.
Cases derived from these datasets keep their citations, their license, and the TCIA data usage policy: https://wiki.cancerimagingarchive.net/x/c4hF.
Pancreatic-CT-CBCT-SEG
Hong, J., Reyngold, M., Crane, C., Cuaron, J., Hajj, C., Mann, J., Zinovoy, M., Yorke, E., LoCastro, E., Apte, A. P., & Mageras, G. (2021). Breath-hold CT and cone-beam CT images with expert manual organ-at-risk segmentations from radiation treatments of locally advanced pancreatic cancer [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/TCIA.ESHQ-4D90. License: CC BY 4.0.
Han, X., Hong, J., Reyngold, M., Crane, C., Cuaron, J., Hajj, C., Mann, J., Zinovoy, M., Greer, H., Yorke, E., Mageras, G., & Niethammer, M. (2021). Deep-learning-based image registration and automatic segmentation of organs-at-risk in cone-beam CT scans from high-dose radiation treatment of pancreatic cancer. Medical Physics, 48(6), 3084-3095. https://doi.org/10.1002/mp.14906
Abdomen cases, cone-beam CT cases and the real scan pairs of Registration Match come from this collection, with the TCIA data usage policy: https://wiki.cancerimagingarchive.net/x/c4hF. Contours of two observers appear as Reader 1 and Reader 2, by number only. Where a cone-beam scan is lined up with a planning CT, the alignment is a computed registration made for Ganform, not the dataset's own.
Colorectal-Liver-Metastases
Simpson, A. L., Peoples, J., Creasy, J. M., Fichtinger, G., Gangai, N., Lasso, A., Keshava Murthy, K. N., Shia, J., D'Angelica, M. I., & Do, R. K. G. (2023). Preoperative CT and Survival Data for Patients Undergoing Resection of Colorectal Liver Metastases (Colorectal-Liver-Metastases) (Version 2) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/QXK2-QG03. License: CC BY 4.0.
Simpson, A. L., et al. (2024). Preoperative CT and survival data for patients undergoing resection of colorectal liver metastases. Scientific Data, 11(1). https://doi.org/10.1038/s41597-024-02981-2. The liver and vessel labels were semi-automatically segmented and verified, so results call them a dataset label. Same TCIA data usage policy.
Beam scenes
Beam Puzzle scenes are single axial slices from LCTSC (CC BY 3.0, with the TCIA data usage policy: https://wiki.cancerimagingarchive.net/x/c4hF) and from the TotalSegmentator subset (CC BY 4.0), each with the couch removed and the body outline of that slice. Each scene keeps its own source's citation and license. The targets are placed by the puzzle and are not in the scans. The app does not generate a skin surface, a face, or a 3D volume rendering.
Synthetic geometry
The case builder can generate geometric volumes for tests. Those volumes are not patient data, not a clinical reference, and they are not in this build.