Monte Carlo simulation of the ELEKTA SL25 with BEAMnrc for IMRT : study of the effects of inhomogeneities on the accuracy of the new treatment planning system VARIAN AAA

Sterpin, Edmond;Tomsej, Milan;Reynaert, Nick;Vynckier, Stefaan
(2006) First European Workshop on Monte Carlo Treatment Planning of the European Workgroup on MCTP — Location: Ghent (22.October.2006)

Files

No attached file found for this publication.

Details

Authors
Abstract
Introduction : Monte Carlo simulation calculates photon/electron transport using physical laws and is widely admitted as the most precise calculation tool for delivered doses by linear accelerator. Indeed, Monte Carlo simulation validity is not limited by delivered treatment complexity (small fields, modulated intensity, inhomogeneities, …) in opposition with analytical algorithms where some inaccuracies may occur in low density material, interfaces and in case of a lack of electronic equilibrium. The Monte Carlo calculation is therefore a first choice tool to calculate accurately dose distributions delivered by IMRT in a very inhomogeneous medium, like in the head and neck case. Material and methods : We build our Monte Carlo model using BEAMnrc, which allows to model different parts of the linac head. For our study, we used a phantom composed by two sets of thin slabs of low-density material separated by two thicker slabs with a density close to water. This phantom allows us to study phenomena like loss of lateral electronic equilibrium, build down and “rebuild up”. The measurement was performed inserting films at different positions along depth axis. The depth dose could be then reconstructed. Our treatment planning system is a VARIAN – ECLIPSE station provided with the new “ Anisotropic Analytical Algorithm (AAA)”, which is a 3D pencil beam convolution/superposition. Inhomogeneities are taken into account by scaling the kernels laterally, adapting depth dependant coefficients defining the pencil beam kernel and introducing a history correction kernel to account for loss of lateral electronic equilibrium. Results : We obtain an excellent agreement (dose differences less than 2% of dose maximum) between measurements and MC calculations in water for all field sizes concerning profiles and depth doses. MC calculations and measurements give also similar results in inhomogeneous phantom, even for small fields (2x2 cm²). We can therefore use in the following of the work our model as a reference for the calculation of the dose in a set of CT images of head and neck patients. Moreover, we compare results with the phantom with the ECLIPSE station’s calculations. We saw good accuracy for depth dose and profiles for 6x6 cm² and 10x10 cm² fields (2-3% maximum deviation). AAA is somewhat less accurate for 2x2 cm², specially at interfaces, but provides significant improvement compare to older ECLIPSE algorithms. These results prove that AAA is handling correctly loss of lateral electronic equilibrium. Some discrepancies where found below bone density cylinder inserted in the middle of the phantom, where the algorithm does not predict dose attenuation while MC and measurements do. AAA is also tested for IMRT plans for four patient cases (one ethmoïd, one head and neck and two lung cases). The results are compared to MC calculations, and significant differences were found for the ethmoïd case, specially for the DVH of the PTV.

Citations

  • APA
  • Chicago
  • FWB

Sterpin, E., Tomsej, M., Reynaert, N., & Vynckier, S. (2006). Monte Carlo simulation of the ELEKTA SL25 with BEAMnrc for IMRT : study of the effects of inhomogeneities on the accuracy of the new treatment planning system VARIAN AAA. First European Workshop on Monte Carlo Treatment Planning of the European Workgroup on MCTP, Ghent. https://hdl.handle.net/2078.5/58070