Treatment planning for MRI guided proton therapy: dose calculations in transverse magnetic fields.
PH-0244
Abstract
Treatment planning for MRI guided proton therapy: dose calculations in transverse magnetic fields.
Authors: Padilla Cabal|, Fatima(1)*[fatima.padillacabal@meduniwien.ac.at];Resch|, Andreas Franz(1);Georg|, Dietmar(1);Fuchs|, Hermann(1);
(1)Medical University of Vienna- Christian Doppler Laboratory for Medical Radiation Research for Radiation Oncology, Department of Radiation Oncology, Vienna, Austria;
Show Affiliations
Hide Affiliations
Purpose or Objective
Magnetic resonance (MR) image guidance is expected to enhance the clinical potentials of particle therapy. Among the several hurdles of integrating an MR in a clinical particle beam line, the current treatment planning strategies are severely challenged. Particles are deflected in magnetic fields, modifying beam paths towards and within the patient. Compensation methods and new dose calculation engines are therefore required. This work aims to develop a research treatment planning system for proton beams in transverse magnetic fields.
Material and Methods
An in-house developed pencil beam (PB) algorithm for MR guided proton therapy (MRgPT) was implemented in the open source matRad research treatment planning system (TPS). The TPS was then utilized to generate treatment plans (TP) in homogeneous phantoms and three patient cases (prostate, liver and brain). The PB model was based on GATE/Geant4 Monte Carlo (MC) simulations in the clinical proton energy range (62.4 – 215.7 MeV) for a dipole magnet and field strengths between 0 and 1 T. The new dose calculation engine in matRad was benchmarked against MC simulations and dose measurements for both single beams and spread-out Bragg peaks (SOBP). Absolute dose was measured with a Roos chamber (PTW, Freiburg, Germany) in a water phantom for magnetic field strengths 0 and 1T. Finally, TPs generated for magnetic field strengths of 0 and 1T were compared. Dose volume histograms (DVH), plan quality indicators (D2%, D50%,D90%,V95%and V105%) and the difference between dose distributions using a global gamma index criteria of 2%/2mm were selected as assessment parameters.
Results
Dose calculations in water of the TPS agreed well with MC simulations, showing gamma index pass rates higher than 99% and 96% for single beams and SOBP, respectively. Deviations between measured, simulated and calculated depth dose distributions were smaller than 2% in the Bragg peak regions for both magnetic field strengths. From a clinical point of view, comparable TPs were obtained for B=0 and B=1T in phantom and patient cases, (see Fig.1-2). Plan quality indicators in the target agreed within 1.5% for all the analysed cases, except for the liver case where deviations up to 2.5% were observed.
Fig. 1 Dose distributions for a prostate patient case for B = 0T (left), B = 1T (middle) and their difference (right). Results are normalised to a prescribed dose to the PTV of 68 Gy.
Fig.2 DVH for the TG119 phantom, prostate, brain and liver patients. Results are displayed for B = 0T (solid lines) and B = 1T (dashed lines).
Conclusion
The in-house developed PB model implemented in the research TPS matRad provides highly accurate dose distributions for protons in transverse magnetic fields for homogenous phantoms and exemplary treatments. More complex benchmarking studies are foreseen to identify the limits of applicability of the proposed model.