S2194
Physics - Inter-fraction motion management and daily adaptive radiotherapy
ESTRO 2026
source and publicly available on GitHub: https://github.com/medical-physics- usz/ARTEMIS
In: Radiotherapy and Oncology 105.2 (2012-11), pp. 220–225. ISSN: 01678140. DOI: 10.1016/j.radonc.2012.08.012Technical Specifications of Radiotherapy Equipment for Cancer Treatment. en. 1st ed. Geneva: World Health Organization, 2021. ISBN: 978-92-4-001998-0Dominique L. Rash et al. “Clinical Response of Pelvic and Para-aortic Lymphadenopathy to a Radiation Boost in the Definitive Management of Locally Advanced Cervical Cancer”. en. In: International Journal of Radiation Oncology* Biology* Physics 87.2 (2013-10), pp. 317– 322. ISSN: 03603016. Keywords: gynae, set-up, rotation From clicks to checks: scripting MR-enhanced online adaptive RT on a conventional C-arm linac. Hubert S Gabry ś , Silvia Fabiano, Riccardo Dal Bello, Sebastian M Christ, Lotte Wilke, Astrid Heusel, Serena Psoroulas, Klara Kefer, Michael Baumgartl, Matthias Guckenberger, Stephanie Tanadini-Lang Department of Radiation Oncology, University Hospital Zurich, Zurich, Switzerland Purpose/Objective: Daily online adaptive radiotherapy using a standalone MRI scanner and treatment delivery at a conventional C-arm linac has the potential to increase the efficiency of treatment but lacks the integration of separate hard- and software solutions. We here evaluated whether Python and Eclipse Scripting API (ESAPI) automation improves workflow reliability and usability. Material/Methods: This daily MRI-enhanced workflow combines a Siemens MAGNETOM Free.Max 0.55 T, Varian C-arm linacs, and sCT generation system from Spectronic Medical. The online adaptive radiotherapy workflow (ARTEMIS) involves: 1) image processing and registration, 2) rule-based plan creation, 3) automatic generation of planning help structures, 4) plan Digital Poster 4948 reoptimization, and 5) plan quality checks. An ESAPI launcher starts a Python application that processes and registers daily T2-weighted MR images, creates a DICOM registration file, and sends the images, registration, and structures copied from the base plan to Eclipse using the DICOM networking protocol. In Eclipse, an ESAPI script clones beam setup and optimization objectives from the base plan, adjusts beam apertures, and generates all needed helper structures. After plan reoptimization, the ESAPI script displays a summary of a comparison of the new plan to the base plan in terms of target size change, monitor unit change, dose per fraction, and reference point equivalence. Our Python and ESAPI code is open-
Results: Implementation of the described automations allowed to significantly reduce the number of individual manual steps in the adaptation process from 13 to 3. As a result, after delivering more than 300 fractions to 66 patients, the median time from imaging start to the plan approval was 34 minutes (IQR, 29-39 minutes). Plan quality and detailed time outcomes are reported in companion submissions. Automation effectively mitigated the risks identified in our initial failure mode and effects analysis (FMEA) of the manual workflow, lowering the highest risk score from 192 points to 24 points. Conclusion: Automation using Python and ESAPI enabled short adaptation times, reduced the risk of manual errors, and lowered cognitive load, allowing us to focus on verification tasks. Additional automation for organs-of- interest autosegmentation and PTV definition could further reduce human-related risks and shorten adaptation times. Keywords: adaptive, scripting, automation
Made with FlippingBook - Share PDF online