MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

Real-Time Online Adaptive Fully Automated Radiation Treatment Re-Planning in Lung Cancer

R A Cormack*, R H Mak , C V Guthier, Department of Radiation Oncology, Brigham and Women's Hospital and Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA

Presentations

(Tuesday, 7/16/2019) 1:45 PM - 3:45 PM

Room: Stars at Night Ballroom 1

Purpose: Anatomical changes during a course of radiation therapy potentially yield a clinically significant increase in dose to normal tissues. This can be mitigated by adaptive strategies which are currently time and resources expensive processes, including an additional planning CT on the day of re-planning. To overcome this limitation, we developed and validated a real-time fully automated plan adaption methodology.Materials and

Methods: The adaptive planning is based on the daily cone-beam CT from which a synthetic-CT (sCT) is generated. The sCT is used to estimate the clinical benefits of a re-plan via a local search algorithm that tailors the treatment plan to the observed anatomical changes. If deemed clinically relevant, a multi-criteria optimization is initiated where a greedy heuristic navigates along the pareto surface in an autonomous way to find the final treatment plan. To allow a seamless workflow the developed algorithms are integrated into our treatment planning system via an application programming interface. The proposed approach was retrospectively validated under IRB approval and includes 20 patients that were previously treated with adaptive radiation therapy. The focus of this study is to validate the proposed approach and to validate if fully automated planning is feasible.

Results: All automatically generated plans were deemed clinically acceptable. The total time for plan generation was 8-20 minutes on a single workstation. A Wilcoxon rank-sum test showed that no statistically significant difference (p>0.05) of target coverage and organ sparing. The mean deviation between key metrics was found to be (0.5±3.7)%.

Conclusion: The proposed approach allows fast online adaptive re-planning without the need of additional planning CTs. The proposed algorithms are capable of efficiently generating complex treatment plans in an unsupervised manner. A fast adaption can significantly decrease the dose to surrounding normal tissues and potentially treatment related side effects.

Keywords

Lung, Optimization, Stereotactic Radiosurgery

Taxonomy

TH- External beam- photons: adaptive therapy

Contact Email