Phase-Contrast Medical Imaging Simulation Using The Monte Carlo Geant4 Code: Evaluation And Optimization Of Physical Processes

Abstract

In this study, phase-contrast medical imaging simulation was evaluated and developed using the Monte Carlo Geant4 code. The primary objective of this research was to develop a simulation tool capable of generating high-contrast images of voxel-based phantoms. For this purpose, the phase-contrast imaging physical process code was designed and implemented, and a phantom with dimensions of 50×50×50 voxels and a resolution of 0.1 mm was selected. The simulation results demonstrated that the external boundaries of the phantom were detected with high accuracy, producing a clear image of these boundaries with light intensity around 51,000 units and contrast fluctuations in the range of 10-15%. Additionally, two significant drops in light intensity (ranging from 7-10%) were observed at distances of 0.005 mm and 0.015 mm, indicating the interference of light rays at these points. As expected, the simulation results successfully identified the external boundaries of the phantom, and in future steps, the focus can be placed on developing the physical process code to enable the detection of internal structures of the phantom. To improve this simulation, it was suggested to employ advanced image processing and signal analysis techniques. Furthermore, using optimization methods based on artificial intelligence could enhance the simulation's accuracy and efficiency. Ultimately, this research demonstrated that the simulation developed using the Monte Carlo Geant4 code holds great potential for advancing medical imaging tools based on phase contrast, although further improvements are needed for internal structure detection and complex light interferences.