Development and validation of radiochromic film dosimetry and Monte Carlo simulation tools for acquisition of absolute, high-spatial resolution longitudinal dose distributions in ocular proton therapy

摘要

Dosimetric validation of prescribed clinical plans in ocular proton therapy requires the use of adapted instrumentation providing high accuracy and spatial resolution. The aim of this work was to develop two independent tools based on Monte Carlo (MC) and EBT3 films to perform high spatial resolution longitudinal dose maps in water for clinical Spread-Out Bragg peaks (SOBP). The first tool is based on EBT3 films, irradiated in beam alignment with a tilt angle of 7° and scanned with high resolution of 300 dpi. Dedicated software was developed to convert Pixel Values (PV) to absolute dose, which makes use of a calibration curve and a correction function of the LET based on MC. MC tool was implemented to generate dose maps in a grid plane with the same configuration (300 dpi, longitudinal alignment, tilt angle of 7°). Both tools were tested for a same SOBP and two collimators of different shapes and apertures. Validation of EBT3 and MC methods was performed by dose map comparison (gamma index test) to a silicon diode and film, respectively. Good agreement was found between diode and film (gamma index >99%) and between MC and film (gamma index >83%). At the plateau, accuracy in dose is better than 0.3% and maximum divergences are of 5% at the SOBP entrance, for both tools. Inconvenience of films is high pixel dispersion (2.3%) and uncertainty in film positioning. MCNPX overestimates angular deflection of Multiple Coulomb Scattering (MCS) and involves lateral profiles 0.5 mmwider at 20% isodose. The mesh tally size in the transversal direction has an impact on dose distributions if spherical or cylindrical beam modifiers are used due to dose convolution in a larger mesh volume, reflected in a slight underdose at lateral edges in deeper depths. The defaults of EBT3 and MC tools concern localized areas outside the Planning Target Volume (PTV) and therefore could potentially be used to perform dosimetric validation of prescribed clinical plans. Work is in progress to further explore their potential and limitations.