Monte Carlo simulation of small electron fields collimated by the integrated photon MLC

• Published in: Physics in medicine & biology 2011-02, Vol.56 (3), p.829-843
• Main Authors: Mihaljevic, Josip, Soukup, Martin, Dohm, Oliver, Alber, Markus
• Format: Article
• Language: English
• Subjects: Electrons; Monte Carlo Method; Photons; Radiotherapy Dosage; Radiotherapy Planning, Computer-Assisted; Radiotherapy, Intensity-Modulated
• ISSN: 0031-9155; 1361-6560
• DOI: 10.1088/0031-9155/56/3/018

In this study, a Monte Carlo (MC)-based beam model for an ELEKTA linear accelerator was established. The beam model is based on the EGSnrc Monte Carlo code, whereby electron beams with nominal energies of 10, 12 and 15 MeV were considered. For collimation of the electron beam, only the integrated photon multi-leaf-collimators (MLCs) were used. No additional secondary or tertiary add-ons like applicators, cutouts or dedicated electron MLCs were included. The source parameters of the initial electron beam were derived semi-automatically from measurements of depth-dose curves and lateral profiles in a water phantom. A routine to determine the initial electron energy spectra was developed which fits a Gaussian spectrum to the most prominent features of depth-dose curves. The comparisons of calculated and measured depth-dose curves demonstrated agreement within 1%/1 mm. The source divergence angle of initial electrons was fitted to lateral dose profiles beyond the range of electrons, where the imparted dose is mainly due to bremsstrahlung produced in the scattering foils. For accurate modelling of narrow beam segments, the influence of air density on dose calculation was studied. The air density for simulations was adjusted to local values (433 m above sea level) and compared with the standard air supplied by the ICRU data set. The results indicate that the air density is an influential parameter for dose calculations. Furthermore, the default value of the BEAMnrc parameter 'skin depth' for the boundary crossing algorithm was found to be inadequate for the modelling of small electron fields. A higher value for this parameter eliminated discrepancies in too broad dose profiles and an increased dose along the central axis. The beam model was validated with measurements, whereby an agreement mostly within 3%/3 mm was found.