Combining Wave-Optics and Monte Carlo Methods for the Simulation of Phase-Sensitive X-Ray Imaging
S Peter1,2*, P Modregger1,3, M K Fix4, W Volken4, P Manser4, M Stampanoni1,2, (1) Swiss Light Source, Paul Scherrer Institut, Villigen, Switzerland, (2) Institute of Biomedical Engineering, ETH Zurich, Switzerland, (3) School of Medicine and Biology, University of Lausanne, Switzerland, (4) Division of Medical Radiation Physics and Department of Radiation Oncology, Inselspital, Bern University Hospital, and University of Bern, SwitzerlandTH-A-141-1 Thursday 8:00AM - 9:55AM Room: 141
Purpose: Due to its high sensitivity towards electron density variations phase-sensitive X-ray imaging is well suited for imaging soft tissue matter. A recently established method is grating interferometry (GI) which produces three complementary types of contrast: absorption, phase and dark-field. However, there are still open questions, for instance about details of the dark-field contrast formation process. This work develops a numerical simulation framework in order to investigate the contrast formation process.
Methods: The framework combines Monte Carlo methods (MC) with wave-optics, since in realistic simulations of phase sensitive X-ray imaging both particle- and wave-like properties of X-rays have to be considered. The source and sample part of the simulation was implemented using the MC transport code egs++. To account for wave-like properties the transport code was expanded to include refraction at surfaces and the optical path length. The gratings and the detector are implemented using wave-optics. The MC part of the framework is connected to the wave-optics part by treating each photon as a plane wave with the same propagation direction. Thus, the data from the MC part can be transformed into a complex amplitude by coherently summing up all plane-waves. The framework is validated by comparing reconstructed image data sets from measured and calculated projections of known phantoms.
Results: The combination of MC and wave-optics has been successfully implemented. The validation of the framework showed very good agreement between simulations and experimental results for all three contrast types. Image correlation coefficients above 0.92 have been found, establishing the framework as a reliable method to model GI.
Conclusion: The developed comprehensive simulation framework, taking refraction and scattering such as Compton and Rayleigh into account, can now be used for detailed investigations of the phase contrast imaging formation process and is particularly suited to study the scattering/dark-field contribution of such processes.