S2319
Physics - Machine learning and AI algorithms
ESTRO 2026
Digital Poster 4978 Reproducible cancer science utilizing the Cancer Genomics Cloud Aditya P Apte 1 , Maria Thor 1 , Aditi Iyer 1 , Eve LoCastro 1 , Viktor Rogowski 2 , Per Munck af Rosenschöld 2,3 , Joseph O Deasy 1 1 Medical Physics, Memorial Sloan Kettering Cancer Center, New York, USA. 2 Medical Radiation Physics, Lund University, Lund, Sweden. 3 Radiation Physics, Skåne University Hospital, Lund, Sweden Purpose/Objective: This work presents a resource for reproducible, out-of- the-box application of analysis workflows for pre- trained radiological and radiotherapy AI models. Analyses are accessible as apps distributed via our public project on the Cancer Genomics Cloud (CGC) platform; the first and only such project supporting CGC, an NCI-funded resource, allows users to collaborate on projects hosted in the cloud by providing tools for managing users and hardware from providers such as AWS and GCP. Access to over 3 Petabytes of publicly available data from the Cancer Research Data Commons (CRDC) ecosystem and repositories such as TCGA and TCIA is provided. Analyses pipelines, distributed as apps, consist of a Docker container with software dependencies, pre- configured hardware requirements, input/output types and scripts for inference. Apps can be run interactively from a web browser or a client computer using the API. Multiple apps can be chained together, with outputs of one app input to the next, to build analysis workflows. Python-native Computational Environment for Radiological Research (pyCERR) is used for data transformation, radiomics and dosimetric computations and visualization. We have deployed clinically validated AI segmentation of over 70 organs utilizing pyCERR utilities. PyCERR radiotherapy images. Material/Methods: additionally provides a user-friendly data structure including simplified access to metadata from various radiological image modalities for downstream analysis. AI inference from radiological images presented here can be readily combined with other data types such as genomics and proteomics from CGC projects. Results: Our public project for AI analyses of radiological and radiotherapy images is available at https://cgc.sbgenomics.com/u/sevenbridges/pycerr- radiological-radiotherapy-imageanalysis. Currently, it includes apps for H&N OAR segmentation using 2D architectures and a foundational 3D SMIT model for thoracic and H&N OARs and lung nodules, trained in- house. We demonstrate the use of our H&N OAR
https://doi.org/10.5152/dir.2019.19321Ma, J et al. (2025). MedSAM2.https://doi.org/10.48550/arXiv.2504.03600v an Griethuysen et al. (2017). PyRadiomics. https://doi.org/10.1158/0008-5472.CAN-17-0339 Keywords: AI Ready, Data Harmonization, Preprocessing Proffered Paper 4953 A Local LLM Agent for Guideline-Based Personalized Prostate Cancer Decision Support in Radiation Oncology – Ready for Clinical Use? Philipp Schubert 1,2 , Annette Schwarz 1,2 , Stefan Knippen 3 , Soeren Schnellhardt 4 , Simon Trommer 5 , Sonia Drozdz 6 , Peter J Goebell 7,2 , Ahmed Gomaa 1,2 , Ricarda Merten 1,2 , Marlen Haderlein 1,2 , Daniel Höfler 1,2 , Thomas Weissmann 1,2 , Matthias May 8,2 , Benjamin Frey 1,2 , Udo Gaipl 1,2 , Bernd Wullich 7,2 , Rainer Fietkau 1,2 , Christoph Bert 1,2 , Stefanie Corradini 1,2 , Florian Putz 1,2 1 Department of Radiation Oncology, Universitätsklinikum Erlangen, Friedrich-Alexander- Universität Erlangen-Nürnberg, Erlangen, Germany. 2 Bavarian Cancer Research Center, (BZKF), Munich, Germany. 3 Department of Radiation Oncology, Helios Clinics of Schwerin-University Campus of MSH Medical School Hamburg, Schwerin, Germany. 4 Department of Radiotherapy and Radiation Oncology, Saarland University Medical Center, Homburg, Germany. 5 Department of Radiation Oncology, University Hospital Halle (Saale), Halle (Saale), Germany. 6 Department of Radiation Oncology, Friedrich-Schiller- University Jena, Jena, Germany. 7 Division of Urology, Universitätsklinikum Erlangen, Friedrich-Alexander- Universität Erlangen-Nürnberg, Erlangen, Germany. 8 Department of Radiology, University Hospital Erlangen, Friedrich-Alexander-Universität Erlangen- Nürnberg, Erlangen, Germany
Made with FlippingBook - Share PDF online