Program Information
Performance of a Hybrid Monte Carlo-Pencil Beam Dose Algorithm for Proton Therapy
A Barragan Montero1*, K Souris1 , D Sanchez-Parcerisa3 , J Lee1 , E Sterpin1,3 , (1) Universite catholique de Louvain, Belgium, (2) Universidad Complutense de Madrid, Spain, (3) KULeuven, Belgium
Presentations
SU-K-108-13 (Sunday, July 30, 2017) 4:00 PM - 6:00 PM Room: 108
Purpose: Analytical algorithms have limited accuracy, especially when modeling very heterogeneous tumor sites. Thus, Monte Carlo (MC) dose calculation is highly desired in proton therapy treatment planning. However, its clinical implementation remains challenging due to the high computational cost. This work addresses the performance of a hybrid optimizer that combines both MC and pencil beam (PB) dose engines to get the best trade-off between speed and accuracy.
Methods: The hybrid optimization strategy calculates the optimal spot weights (w) using the analytical beamlets matrix (Pᴾᴮ) and a correction term C. After a first optimization where C = 0, the method alternates optimization of w using Pᴾᴮ with updates of C = Dᴹᶜ – Dᴾᴮ, where Dᴹᶜ results from a regular MC computation and Dᴾᴮ = Pᴾᴮ * w. Updates of C can be triggered as often as necessary by calling the MC with the last corrected weights w as input. The hybrid method was applied to three cases (prostate, lung and brain) and compared with full-MC plans (Pᴹᶜ). For simplicity, PTV-based plans were created but the method can be equally applied to robust optimization. The plans were created with MIROpt, our in-house treatment planning system (TPS), which is coupled with a PB algorithm (FoCa) and a super-fast MC (MCsquare).
Results: For all patients, the hybrid method achieved excellent target coverage (D₉₅ > 95% of dose prescription) after only a single update of C. The PTV D₉₅ and D₅, as well as the dose to organs at risk, were equivalent to the full-MC plans. The computation time of hybrid plans was reduced by a factor 7 w.r.t the full MC plans.
Conclusion: This study demonstrated the successful performance of hybrid MC-PB optimization for proton therapy, which can be immediately considered as an option for improving dose calculation accuracy of commercial analytical TPSs.
Contact Email: