An X-ray scatter system for material identification in cluttered objects: A Monte Carlo simulation study


The analysis of X-ray scatter patterns has been demonstrated as an effective method of identifying specific materials in mixed object environments, for both biological and non-biological applications. Here we describe an X-ray scatter imaging system for material identification in cluttered objects and investigate its performance using a large-scale Monte Carlo simulation study of one-thousand objects containing a broad array of materials. The Geant4 Monte Carlo source code for Rayleigh scatter physics was modified to model coherent scatter diffraction in bulk materials based on experimentally measured form factors for 33 materials. The simulation was then used to model coherent scatter signals from a variety of targets and clutter (background) materials in one thousand randomized objects. The resulting scatter images were used to characterize four parameters of the imaging system that affected its ability to identify target materials: (a) the arrangement of materials in the object, (b) clutter attenuation, (c) type of target material, and (d) the X-ray tube current. We found that the positioning of target materials within the object did not significantly affect their detectability; however, a strong negative correlation was observed between the target detectability and the clutter attenuation of the object. The imaging signal was also found to be relatively invariant to increases in X-ray tube current above 1 mAs for most materials considered in the study. This work is the first Monte Carlo study to our knowledge of a large population of cluttered object of an X-ray scatter imaging system for material identification and lays the foundation for large-scale studies of the effectiveness of X-ray scatter imaging systems for material identification in complex samples. © 2014 Elsevier B.V. All rights reserved.