Hybrid Monte Carlo and deterministic simulation approach for modeling a computed radiography imaging chain from X-ray exposure to optical readout

摘要

Simulation of radiographic inspection is of great interest for experimental outcomes prediction and optimal operating condition determination. As concerns computed radiography (CR), the use of photo-stimulable imaging plates and laser scanners, implies modeling the behavior of a multi-stages detector. As a consequence, both the X-ray and the optical system responses have to be handled. Moreover, for high energy X-rays, two issues often trouble CR simulation: long running time and X-ray scattering image contribution, which should not be neglected. To overcome these issues, we have developed a complete hybrid model which is the first available one at such energies. In our approach, the imaging process is decomposed into three independent successive stages: X-ray attenuation by an object, X-ray latent image generation, and optical readout. A deterministic code is applied to obtain rapidly the transmitted X-ray image emerging from a complex object. The energy deposition is then simulated by a convolution of the transmitted X-ray image with a CR detector response model, which was obtained off-line by a Monte Carlo tool. Then, optical readout is modeled using the same hybrid approach, where the optical response (laser light spreading in the imaging plate) was obtained by Monte Carlo and laser scanning is modeled analytically. A good agreement has been observed between the proposed hybrid model and a full Monte Carlo approach for the X-ray energy deposition stage. A realistic X-ray inspection case study has been chosen to emphasize the interest of this complete hybrid model. The comparison of three different detector configurations and the influence of readout laser power are illustrated.