Computational Method to Characterise the SNR-Dose ratio of X-Ray Fluorescence in Complex Biomedical Samples Biomarked for microCT and XFCT

This study proposes an innovative methodology that combines experimental microCT imaging with Monte Carlo irradiation simulations using the PENELOPE code aimed at enabling detailed analyses of different features involved in the irradiation processes along with elemental mapping, as required for X-ray fluorescence imaging of in vivo samples. By examining the relationship between signal-to-noise ratio (SNR) and absorbed dose (D), a mathematical approach was proposed to determine critical points where their rates of change are equal. Thus, allowing to identify optimal concentrations of gold dispersions to the specific simulated system of interest, where the signal quality efficiently improves as weighted in relation to the absorbed dose. The proposed methodology has been applied to a typical small animal (rat) microCT that was further used by the computationally implemented model to infuse the animal kidneys by different amounts of gold. For the K_(α_1 ), K_(α_2 ), and K_(β_1 ) lines, these critical concentrations were found to be 0.78 %, 1.32 %, and 0.32 % w/w, respectively, while no such behavior was identified for the K_(β_2 ) line under the given configuration considered as a representative/typical small animal infused with Au-based agents. The obtained results highlight the methodology's ability to optimize the balance between the absorbed dose and corresponding SNR, obtaining high-quality images without compromising in vivo samples due to excessive radiation exposure. Moreover, the proposed methodology provides high-resolution structural imaging and detailed elemental mapping, facilitating the analysis of detection limits along with the overall system performance. These findings confirm the robustness and reliability of the approach, offering a valuable tool for refining X-ray spectroscopy imaging processes and advancing X-ray fluorescence-based tomography techniques.