MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

On Monte Carlo Deliverable Pareto Surface Navigation for Photon IMRT and Mixed Beam Radiotherapy

S Mueller1*, F Mueller1 , M K Fix1 , S Tessarini1 , T Risse1 , O Elicin1 , K Loessl1 , M F M Stampanoni2 , P Manser1 (1) Division of Medical Radiation Physics and Department of Radiation Oncology, Inselspital, Bern University Hospital, and University of Bern, Switzerland (2) Institute for Biomedical Engineering, ETH Zurich and PSI, Villigen, Switzerland

Presentations

(Sunday, 7/29/2018) 1:00 PM - 1:55 PM

Room: Karl Dean Ballroom A1

Purpose: Treatment plans navigated on an approximated Pareto surface using conventional approaches require subsequent procedures like leaf-sequencing and deliverable dose calculations leading to dose discrepancies between navigated and deliverable dose distributions. To overcome these issues, the aim of this work is to develop a multicriteria optimization supporting Monte Carlo (MC) deliverable Pareto surface navigation applicable to photon IMRT and photon multileaf collimator (pMLC) based mixed photon-electron beam radiotherapy (MBRT).

Methods: A column generation based multicriteria direct aperture optimization is developed, which calculates the MC-deliverable dose distributions of iteratively added apertures during optimization (del-MCDAO), i.e. taking the impact of the pMLC into account. Navigated plans consist of apertures, which are shared over the whole Pareto surface approximation and other apertures, which are locally shared over sub regions of the Pareto surface. The del-MCDAO is applied for a chest wall and a lung case for photon IMRT and MBRT and compared in terms of practicability and performance to the same MCDAO operated in the conventional scheme (con-MCDAO). Conventional means that the deliverable dose calculation is performed after navigation and not during optimization.

Results: For the investigated cases and treatment modalities, the computation time for generating the Pareto surface approximation takes about 5 h longer for del-MCDAO compared to con-MCDAO with 100 CPU-cores. However, each time the MC-deliverable dose distribution of a navigated plan is wanted to be displayed using con-MCDAO during navigation, the planner needs to take up to 1 h of calculation time with 100 CPU-cores into account. These waiting times are circumvented with del-MCDAO. Furthermore, treatment plan quality of navigated plans is similar between del-MCDAO and con-MCDAO.

Conclusion: del-MCDAO successfully shifts computation time from post-navigation procedures into optimization to avoid a human iteration loop. This results in an improved treatment planning workflow. This work was supported by Varian Medical Systems.

Funding Support, Disclosures, and Conflict of Interest: This work was supported by Varian Medical Systems.

Keywords

Optimization, Inverse Planning, Monte Carlo

Taxonomy

TH- External beam- photons: IMRT dose optimization algorithms

Contact Email