Monte Carlo Dose Estimation in Patient-Specific Volume Mesh From CT Images

Radiation treatments are in continuous evolution and the use of increasingly advanced technologies enables personalized treatments. According to this, the work presented is focused on the 3D computed reconstruction of segmented CT images taken from an anthropomorphic phantom (RANDO phantom) using unstructured meshes. This phantom model is used to simulate dose distribution during a beam irradiation with Monte Carlo code. Searching for the most appropriate and rapid methodology for meshing the segmented phantom is the main goal of the present work. The use of different coupled softwares, as image segmentation software or meshing geometry software were necessary. Besides, several FORTRAN routines were coded for conversion formatting to Monte Carlo code and dose calculation. This study has used the last version of MCNP (v.6.1.1) Monte Carlo code to simulate a RANDO phantom irradiation. This new code version presents a new tool to improve geometry description by the importation of meshed models. The generated meshed head model is included in a Monte Carlo model of photon beam transport emitted by a medical LINear ACcelerator (LinAc). Results have been compared with values obtained by experimental measurements in different positions inside the phantom using High-Sensitivity MOSFET (Metal Oxide Semiconductor Field Effect Transistor) dosimeters. Monte Carlo results are in good agreement with experimental values, which demonstrate the validity and accuracy of the Monte Carlo method to estimate dose. The innovation of using meshed geometries in these calculations, shows an important advantage defining complex geometries as human anatomy.