VMC++ versus BEAMnrc: a comparison of simulated linear accelerator heads for photon beams.

Division of Medical Radiation Physics, Insel Hospital, University of Berne, Berne 3010, Switzerland.
Medical Physics (Impact Factor: 3.01). 04/2008; 35(4):1521-31. DOI: 10.1118/1.2885372
Source: PubMed

ABSTRACT BEAMnrc, a code for simulating medical linear accelerators based on EGSnrc, has been bench-marked and used extensively in the scientific literature and is therefore often considered to be the gold standard for Monte Carlo simulations for radiotherapy applications. However, its long computation times make it too slow for the clinical routine and often even for research purposes without a large investment in computing resources. VMC++ is a much faster code thanks to the intensive use of variance reduction techniques and a much faster implementation of the condensed history technique for charged particle transport. A research version of this code is also capable of simulating the full head of linear accelerators operated in photon mode (excluding multileaf collimators, hard and dynamic wedges). In this work, a validation of the full head simulation at 6 and 18 MV is performed, simulating with VMC++ and BEAMnrc the addition of one head component at a time and comparing the resulting phase space files. For the comparison, photon and electron fluence, photon energy fluence, mean energy, and photon spectra are considered. The largest absolute differences are found in the energy fluences. For all the simulations of the different head components, a very good agreement (differences in energy fluences between VMC++ and BEAMnrc <1%) is obtained. Only a particular case at 6 MV shows a somewhat larger energy fluence difference of 1.4%. Dosimetrically, these phase space differences imply an agreement between both codes at the <1% level, making VMC++ head module suitable for full head simulations with considerable gain in efficiency and without loss of accuracy.

1 Follower
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: For superficial lesions, the electrons may be used for radiation therapy. The high energy photons and electrons are produced by a Linear accelerator (Linac). Many of electron fields need the shielding of normal or critical organs. The electron shields are usually lead slabs with few millimeter thicknesses which should be placed near the skin, less than 1 cm away from skin. In the inspection of patients setting in a clinic by a physicist, it was noted that in some cases the technician places the shields far away from skin in the way that the shadow of the field still matches the shielded area. This is due to a conceptual mistake in which one assumes that electrons travel in a straight line and matching the shadow of lead slab is enough for the shielding. This project is about Monte Carlo simulation of this case and dosimetry in which the excess dose to the tissue under the shield is calculated. In this study, BEAMnrc and DOSXYZnrc are used for simulation of the Linac and the electron shields. The water phantom as well as the Linac head (NEPTON Linac) is simulated in the electron mode. The simulation is performed in three various cases in which the lead shield is placed in distances of 1, 20, 40 cm from the surface of the phantom. In all cases, the edge of the shield is matched with the light field, so the shields get smaller as they move from the surface because of the divergence of the light field. The simulations were done in two energies, 6 and 13 MeV. The experiments also were done with EDR2 film dosimetry and the simulation results were validated using the experimental results. In all cases, the dose under the shield was normalized to the dose in the center. The dose of the normal organ under the shield was 2, 38, 43% with respect to the center for shield distances of 1, 20, 40 cm, respectively. So there is a considerable increasing of the dose due to the distanced shielding. In this work exact amount of the dose from this mistake (distanced shielding) is calculated and simulated.
    03/2012; 2(3):144-148.
  • Medical Physics 01/2008; 35(11). DOI:10.1118/1.2998203 · 3.01 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Minimizing the differences between dose distributions calculated at the treatment planning stage and those delivered to the patient is an essential requirement for successful radiotheraphy. Accurate calculation of dose distributions in the treatment planning process is important and can be done only by using a Monte Carlo calculation of particle transport. In this paper, we perform a further validation of our previously developed parallel Monte Carlo electron and photon transport (PMCEPT) code [Kum and Lee, J. Korean Phys. Soc. 47, 716 (2005) and Kim and Kum, J. Korean Phys. Soc. 49, 1640 (2006)] for applications to clinical radiation problems. A linear accelerator, Siemens’ Primus 6 MV, was modeled and commissioned. A thorough validation includes both small fields, closely related to the intensity modulated radiation treatment (IMRT), and large fields. Two-dimensional comparisons with film measurements were also performed. The PMCEPT results, in general, agreed well with the measured data within a maximum error of about 2%. However, considering the experimental errors, the PMCEPT results can provide the gold standard of dose distributions for radiotherapy. The computing time was also much faster, compared to that needed for experiments, although it is still a bottleneck for direct applications to the daily routine treatment planning procedure.
    Journal- Korean Physical Society 05/2012; 60(9). DOI:10.3938/jkps.60.1433 · 0.43 Impact Factor


Available from