Abstract

In this work we describe and simulate a setup which allows the X-ray Fluorescence (XRF) detection of gold nanoparticles (GNPs) excited by a convergent orthovoltage X-ray source, immersed in a big matrix of light elements (water). Under these conditions the problem is the poor Signal-to-Noise Ratio (SNR), which under normal irradiation conditions limit the detection of the GNPs. For the X-ray source (Teledyne CP120B) multiple calculations using SpekCalc software were run to determine an optimum external filtering thickness and collimation material, where was found that 1 mm of Cu external filtering improves the ratio and quantity of useful photons above the Au K-edge absorption energy. Monte Carlo (PENELOPE v2008) simulations were performed using the confocal geometry. The scenarios, including the composition (lead and bronze alloys), shape (cylindrical, conical) and location of the detectors were defined. Using a set of combinations for collimators in the X-ray source and detectors, where both are aligned to a unique focal point, it was possible to increase the SNR of the XRF signal with respect to Compton radiation in a specific region within the phantom, which makes possible to identify the location of the GNPs. The best SNR (185%) was achieved using a 4 cm length conical collimator made of lead. This would allow the development of a new technique for functional medical imaging, and according with the outcome of these simulations, in subsequent stages a prototype will be developed to validate the simulation outcomes, and confirm that it is possible to detect the location of GNPs from a human scale phantom based on their fluorescent emission.