TH-AB-204-11: X-Ray Fluorescence CT Induced by Proton Beam: Experiments and Simulations

Abstract

Purpose: To demonstrate the feasibility of x-ray fluorescence computed tomography induced with proton beams (pXFCT) for imaging of gold contrast agent by means of experiments and Monte Carlo (MC) simulations. Methods: A 7-cm diameter water phantom containing 2.2-cm diameter vials filled with gold solutions of 3–5% Au (percent weight concentration) was imaged with pXFCT using a 7-mm FWHM 220-MeV proton beam and a 3×3mm2 CdTe photon-counting detector. The phantom was imaged in 1st generation CT scanner geometry using a programmable rotation/translation stage and 21 translation steps separated by 3.3 mm and 36 rotation steps in 10° intervals. Each of the 756 x-ray spectra was acquired for 20 s using 5×101⁰ incident protons with the CdTe detector placed at 45 cm from the isocenter and at 90° with respect to the proton beam. The 220 MeV proton beam was stopped in a solid water beam dump and the total imaging time was 4.2 hours. The experimental pXFCT data acquisition geometry was modeled based on the actual and a simplified geometry with the TOPAS MC code. pXFCT images were reconstructed based on experimental and MC-simulated x-ray spectra with filtered back-projection using Kα peaks of gold. Results: All gold vials were visible in both the experimental and simulated pXFCT images. Contrast-to-noise ratio (CNR) of the 3% Au vial was 5.8 and 11.5 in the experimental and simulated pXFCT image, respectively. pXFCT detection limit of the experimental setup was determined to be 1.8% Au, which was twice as high as the MC-simulated detection limit. Further MC simulations revealed that x-ray scatter from the beam dump was the main contribution to x-ray fluorescence signal contamination. Conclusion: We have demonstrated the feasibility of proton-induced XFCT imaging of gold. We anticipate that pXFCT imaging sensitivity will be improved in an optimized pXFCT imaging system utilizing beam collimation.