Dosimetric Simulation of a Talbot-Lau X-Ray Interferometry System for Medical Diagnostic Radiography

Abstract

Purpose: Lung cancer and other pulmonary disease is typically diagnosed by means of X-ray imaging, such as conventional radiography and computed tomography (CT). However, these modalities are not ideal for lung imaging due to their poor soft tissue contrast and the microscopic nature of early-stage disease. Diagnosis of lung disease could potentially be improved by an imaging modality which is as convenient and low-dose as current X-ray imaging, but provides better visualization of lung tissue. Methods: We are developing chest radiography and CT systems for medical imaging which use a triple-grating Talbot-Lau X-ray interferometry (TLXI) design to extract dark-field and phase contrast information alongside a typical absorption image. The presence of the diffraction gratings and the requirements of the imaging protocol potentially impact patient radiation dose and image quality; assessing these impacts is key to maximizing image quality for minimum radiation dose. Furthermore, the effect on image quality of mechanical strain on the optical components has not been closely studied. We have produced software applications which employ 3D Monte Carlo particle transport techniques and finite element stress analysis (FEA) to calculate the radiation dose delivered to the patient for both radiography and CT and the rotation-induced deformation of the diffraction gratings for CT. A set of test cases were used to valid the software models. These numerical simulations will be used in the future to elucidate the relationship of image quality to acquisition parameters and grating design. Results: The test cases confirmed that the geometry setup and output dosimetry data both meet real-world expectations. Our tests showed that the diagnostic X-ray beam interacts realistically with the diffraction gratings and patient volumes within the simulation geometries. The FEA results also showed that rotational motion within the CT setup results in only small deformations to the structure of the optical components; the FEA results can be coupled to a model of TLXI signal propagation in the future to quantify any impact on image quality. Conclusion: We have demonstrated the validity and functionality of our simulation models. The toolkit is ready for use in assessing prototype design to yield attenuation, dark-field, and phase contrast images of sufficient quality while remaining within acceptable dose tolerances. The addition of dark-field and phase contrast images to absorption images from radiography and CT is expected to enhance diagnosis of early-stage lung cancer and other pulmonary disease.