Fast and accurate Monte Carlo simulation of the Tomotherapy treatment unit in the static mode
(2007) Belgian Hospital Physicists Association Symposium — Location: Liège (9.February.2007)
Files
No attached file found for this publication.
Details
- Authors
- et. al.
- Abstract
- Introduction Tomotherapy treatment unit is designed to deliver highly modulated IMRT treatments. Because of the complexity of the treatment, inaccuracies may occur for convolution/superposition based algorithms specially in inhomogeneities. Monte Carlo (MC) calculations’ validity, instead, is not limited by the complexity of the treatment and it is therefore a tool of choice to calculate accurately dose distribution provided by IMRT treatments, specially in a highly heterogeneous media (head and neck tumors, …). The concept of Tomotherapy provides new challenges in MC simulations, since simultaneous movement of the gantry, the couch and the multi-leaf collimator (MLC) must be simulated accurately. However, before accounting for gantry and couch movement, high accuracy must be achieved while simulating the static mode (gantry and MLC are not moving during irradiation). The aim of this study is to firstly model the static mode, i.e. static open fields as well as the MLC. Since the transport of the particles through the MLC is time consuming, the second purpose of the work is to find a technique where fast simulation of the MLC is performed without losing significant accuracy. This is done through analytical operations thanks to a transformation called “transfer function” (TF). Material and methods During the commissioning process of the MC model, the calculations are compared to measurements performed in a water tank, with an EXTRADIN A1SL ion chamber (Standard Imaging), with the SSD equal to 85 cm. Moreover, for some MLC configurations when high geometric resolution is required, profiles are measured with EDR-2 film in a solid water phantom (SSD = 85 cm, depth = 1.5 cm). The MC simulation is based on the PENELOPE code, which can simulate any geometry configuration, provided that the different elements of the geometry are delimited by quadric surfaces. Since the geometry and materials are fully disclosed by the manufacturer, negligible approximations are made in our MC geometry input file. The only parameters that can be tuned on are therefore electron source spot size and electron energy. The model is commissioned the three static open fields available in Tomotherapy treatment units (1x40cm², 2.5x40cm² and 5x40cm²). The MLC is also defined according to detailed technical drawings. The validation of the geometry is performed thanks to MLC configurations where it is possible to verify the geometry of each individual leaves as well as the whole leaf bank. TF is applied as follows. For all the particles (photons) of the phase space, at most three attenuation coefficients are linked and calculated according to the raylength of the photon in each leaf it passes. If the photon rayline passes through more than three leaves, the particle is killed. In a specific MLC configuration, the weight of all the particles will or will not be multiplied by the attenuation coefficient depending on which leaf is closed or open. It should be mentioned that the attenuation coefficients are calculated using the cross section database of PENELOPE. Results Good agreement is obtained for the three static open fields with deviations smaller than 2% of dose maximum and 1 mm in high dose gradients region. Similar results were achieved for the MLC in the full MC mode. With the TF, it was shown that no significant deviations were introduced, except in the build up region were the absence of contamination electrons in the transformed phase space file yields to some differences. Thanks to the tools developed in this study, a reliable and fast simulation of Tomotherapy in the helical mode can now be implemented.
- Affiliations
Citations
Sterpin, E., Vynckier, S., & et al. (2007). Fast and accurate Monte Carlo simulation of the Tomotherapy treatment unit in the static mode. Belgian Hospital Physicists Association Symposium, Liège. https://hdl.handle.net/2078.5/58797