Monte-Carlo simulation for the optimization of a fan-beam X-ray backscatter system for the identification of human remains

This study presents a comprehensive simulation and optimization of a ground-penetrating X-ray (GPX) backscatter system using the Geant4 Application for Tomographic Emission (GATE)-based Monte-Carlo method. Our research focuses on enhancing fan-beam-based GPX inspection systems for the detection of buried objects, particularly in the context of recovering Korean War fallen soldiers' remains. We employed both pixel and depth collimators to improve depth selectivity and image contrast in challenging subsurface environments. The simulations were conducted using a fan-beam X-ray source with a tube voltage of 150 kVp and a tube current of 50 μAs. While actual field operations may require mA-level currents, computational constraints limited our simulations to this lower, yet reasonably achievable, photon count. We investigated the system's performance across various soil densities (0.2 to 1.0 g/cm³) and burial depths (up to 90 mm). The results demonstrated that the ability to detect objects buried up to 70 mm deep under specific soil conditions, with stone exhibiting the highest detection potential compared to polyethylene and bone. Our findings highlight the significant impact of soil density on image quality and detection limits. We established a relationship between X-ray energy, penetration depth, and signal attenuation, providing valuable guidance for optimizing X-ray source parameters for deeper object detection. This research represents the first optimization of a ground-penetrating X-ray backscatter system specifically leveraging the GATE-based Monte Carlo simulation framework. The gained insights offer a robust foundation for enhancing GPX technology, potentially leading to significant advancements in subsurface imaging and detection methodologies for applications in humanitarian demining and forensic archaeology. • First GPX backscatter system optimized via GATE Monte-Carlo simulations . • Improved depth selectivity and contrast using fan-beam collimator designs . • Object detection shown at 70 mm depth under optimal soil conditions . • Adaptive imaging proposed for varying soil densities in real scenarios. • Potential applications in humanitarian demining , offering a non-invasive solution with minimal environmental impact.