Volume 3 - No 5 - Special Edition: Technological Innovation
DEVELOPMENT OF A NONPARAMETRIC MONTE CARLO METHOD FOR SIMULATION OF PLANE SOURCES OF RADIATION IN EXPOSURE COMPUTER MODELS
In exposure computational models (ECMs) of DEN/UFPE an anthropomorphic phantom is coupled to EGSnrc Monte Carlo (MC) Code to perform dosimetric estimates. The models also need a simulator of the emission source of ionizing radiation. This paper presents a nonparametric MC algorithm of a plane source of radiation, isotropic for rotation, dependent on the radial variable and the cumulative distribution function (CDF) of the problem, known but not invertible. The algorithm chooses the N values of the variable radial and calculates the corresponding values of the probability density function (PDF) and FDA of the problem. It is supposed that, if the CDF values were discrete, they correspond to radius frequencies (RF) in consecutive intervals of the domain. Using a random number generator (RNG) uniform normalized in [0, 1), the MC technique of generalized inversion, and linear interpolation it was possible to obtain samples of r. For dosimetric comparisons, MSTA (Mash STAnding = MASH phantom + EGSnrc + algorithm based on exponential PDF) ECM was reused and MSTA_RF ECM was built. The implementations were added to the MonteCarlo software, developed by the authors and updated whenever necessary. The results presented and commented establish a new algorithm for a plane source of radiation.
PaperID: p 234-250
Author's Name: José Wilson Vieira, Viriato Leal Neto, José de Melo Lima Filho and Fernando Roberto de Andrade Lima
Volume: Volume 3
Issues: No 5 - Special Edition: Technological Innovation
Keywords: Exposure computational models, plane sources of radiation, nonparametric Monte Carlo methods.