Development and machine learning of a multi-criteria ionizing radiation dose distribution model in the Eclipse treatment planning system

The automation of the radiotherapy preparation process is demonstrated through the development and machine learning of a multi-criteria ionizing radiation dose distribution model, using artificial intelligence tools embedded in the RapidPlan module of the Eclipse v16.1 (Varian Medical Systems) treatment planning system. A retrospective data analysis of 40 patients with thoracic and lumbar spine pathologies was performed to train the model. For each patient, a radiation dose distribution model was created using stereotactic radiation therapy with an inverse planning method and a dose fractionation regimen of 6 Gy in 5 fractions. The performance of the developed model was evaluated on a test set of 10 patients. Verification results confirm the model’s suitability for clinical application in oncological healthcare facilities and the prospect of using it to create personalized treatment plans. Automation of the pre-radiotherapy preparation process reduced the time spent on computer modeling of the three-dimensional ionizing radiation dose distribution and improved the quality of specialized medical care provided by stereotactic radiation therapy.

1. The efficacy of external beam radiotherapy and stereotactic body radiotherapy for painful spinal metastases from renal cell carcinoma / G. K. Hunter, E. H. Balagamwala, S. A. Koyfman [et al.] // Practical Radiation Oncology. – 2012. – Vol. 2, № 4. – P. 95–100. https://doi.org/10.1016/j.prro.2012.01.005

2. Stereotactic body radiotherapy for metastatic spinal sarcoma: a detailed patterns-of-failure study / J. E. Leeman, M. Bilsky, I. Laufer [et al.] // Journal of Neurosurgery. – 2016. – Vol. 25, № 1. – P. 52–58. https://doi.org/10.3171/2015.11.SPINE151059

3. Reducing inter- and intra-planner variability in radiotherapy plan output with a commercial knowledge-based planning solution / A. Scaggion, M. Fusella, A. Roggio [et al.] // European Journal of Medical Physics. – 2018. – Vol. 53. – P. 86–93. https://doi.org/10.1016/j.ejmp.2018.08.016

4. Eclipse Photon and Electron Algorithms Reference Guide. – Varian Medical Systems, Inc., 2014. – Ch. 10: DVH Estimation Algorithm for RapidPlan. – P. 219–229. – URL: https://jpneylon.github.io/ABR/PDFs/Add_052418/ EclipseAlgorithms13.6_RefGuide.pdf (date of access: 01.02.2025).

5. Искусственный интеллект в медицинской физике: функции, обязанности, образование и подготовка медицинских физиков клинической квалификации / IAEA (Междунар. агентство по атом. энергии). – Вена, 2025. – 43 с. – URL: https://www-pub.iaea.org/MTCD/Publications/PDF/TCS-83_R_web.pdf (дата обращения: 01.02.2025).

6. Normal tissue constraints for SRS/SBRT. – URL: https://www.iba-dosimetry.com/fileadmin/user_upload/products/02_radiation_therapy/myqa_srs/iba_dose_contraints_poster_rev2_0522.pdf (date of access: 01.02.2025).

7. Machine learning and modeling: data, validation, communication challenges / I. El Naqa, D. Ruan, G. Valdes [et al.] // Medical Physics. – 2018. – Vol. 45, Iss. 10. – P. 834–840. https://doi.org/10.1002/mp.12811

8. Effect of Dosimetric Outliers on the Performance of a Commercial Knowledge-Based Planning Solution / A. R. Delaney, J. P. Tol, M. Dahele [et al.] // International Journal of Radiation Oncology Biology Physics. – 2016. – Vol. 94, Iss. 3. – P. 469–477. https://doi.org/10.1016/j.ijrobp.2015.11.011

9. Development and evaluation of a clinical model for lung cancer patients using stereotactic body radiotherapy (SBRT) within a knowledge-based algorithm for treatment planning / K. C. Snyder, J. Kim, A. Reding [et al.] // Journal of Applied Clinical Medical Physics. – 2016. – Vol. 17, Iss. 6. – P. 263–275. https://doi.org/10.1120/jacmp.v17i6.6429

10. An analysis of knowledge-based planning for stereotactic body radiation therapy of the spine / J. J. Foy, R. Marsh, R. K. Ten Haken [et al.] // Practical Radiation Oncology. – 2017. – Vol. 7, Iss. 5. – P. E355–E360. https://doi.org/10.1016/j.prro.2017.02.007

11. Can knowledgebased DVH predictions be used for automated, individualized quality assurance of radiotherapy treatment plans? / J. P. Tol, M. Dahele, A. R. Delaney [et al.] // Radiation Oncology. – 2015. – Vol. 10, № 1. – Art. ID 234. https://doi.org/10.1186/s13014-015-0542-1