Characterization and Monte Carlo Simulation of Low- and High-Perturbation In-Vivo Diode Dosimeters for 9 Mv X-Rays Publisher



Moslehshirazi MA¹ ; Shahbazigahrouei D² ; Karbasi S² ; Monadi S² Show Affiliations
Source: IFMBE Proceedings Published:2009

Abstract

The use of semiconductor diode dosimeters is a recommended and practical method of in-vivo dosimetry for radiotherapy treatment verification. The dependence of diodes' sensitivity on various factors necessitates the application of correction and calibration factors. There is very little published data on such factors for low-perturbation (thin-buildup) diodes and no data on diode characterization for 9 MV photons. This study aimed to compare the correction factors for SSD, field size, temperature and obliquity of two types of Scanditronix/IBA diodes (EDP10 and EDD5, with full and thin buildup, respectively) in 9 MV x-rays, and design reliable MCNP Monte Carlo models for these diodes. The AAPM protocol was followed for acceptance testing, calibration and measurement of correction factors. The measurements and simulations were made with the diodes positioned on the horizontal surface of a polystyrene phantom. Temperature correction factors of 0.29%/°C for EDP10 and 0.27%/°C for EDD5 were measured for temperature range 19-38°C. The percentage variation of other correction factors were as follows (EDP10, EDD5): SSD range 80-120 cm (2.6%, 3.8%); field size range 5×5 to 40×40 cm2 (4.2%, 21.4%); axial-angle obliquity range -80° to 80° (4.0%, 17.6%); tilt-angle obliquity from 0° to 80° (center to head of diode) (2.0%, 19.0%); tilt-angle obliquity from 0° to -80° (center to tail of diode) (4.4%, 31.0%). Almost all of the simulation results were within 4% of the measurements. Although the low-perturbation EDD5 diode can be used for accurate in-vivo dosimetry, its CFs are much larger than the EDP10 and, therefore, should be applied carefully. The results indicate against the use of the EDD5 in axial angles beyond ±50° and exceeding the range 0° to +50° tilt angle. Having obtained reasonably good agreement between simulations and measurements, the model may be used to perform further studies, especially where measurements are difficult, inaccurate or impossible. © 2009 Springer-Verlag.