Sensor-Based X-Ray Phase-Contrast and Dark-Field Imaging with a 1 μm Detector

Sammanfattning

X-ray imaging is a widely used diagnostic tool in medicine, traditionally relying on attenuation contrast to differentiate between tissues. However, alternative imaging modalities such as phase-contrast and dark-field imaging offer additional contrast mechanisms that are able to improve the detection of soft tissue structures. This thesis investigates the potential of sensor-based X-ray phase-contrast and dark-field imaging using a high-resolution detector with a resolution of 1 μm.A simulation framework was developed to model the propagation of a monochromatic, parallel X-ray wavefront through different sample geometries. Both dark-field and phase-contrast signals were analysed independently under idealised conditions. The dark-field signal was studied first using simplified systems of randomly distributed sil- ica spheres, allowing for validation against analytical models and experimental data. Subsequently, the framework was applied to more complex biological structures, specifically human skull bone, to assess its effect on the dark-field signal and its potential impact on brain imaging.In addition, the capability of phase-contrast imaging to improve the detectability of small cancerous lesions in soft tissue was evaluated. A comparison with conventional attenuation-based imaging was performed using the contrast-to-noise ratio (CNR) as a metric, taking into account feature size and radiation dose.The results demonstrate good agreement between simulation, analytical predictions, and experimental findings for the dark-field signal in model systems. Simulations of skull bone indicate that its microstructure can generate a substantial dark-field signal, which may interfere with imaging of underlying soft tissue, but must be tested in an experimental setting before conclusions can be drawn. Furthermore, phase-contrast imaging shows improved detectability of small structures compared to attenuation- based imaging, particularly when using high-resolution detectors.Overall, this work highlights both the potential and the challenges of combining phase- contrast and dark-field imaging in medical applications. The developed simulation framework provides a flexible tool for studying these imaging modalities and may support future optimisation and investigation of imaging systems and reconstruction methods.