Conventional radiography is largely insensitive to early microstructural alterations in lung tissue, and CT—while more informative—carries dose levels that make it unsuitable for widespread screening. Dark‑field imaging fills this gap by measuring ultra‑small‑angle scattering arising from the alveolar network. Early studies demonstrate strong potential but rely on fine‑pitch absorption gratings that are costly, difficult to manufacture, and may reduce attenuation image quality.
This project addresses these limitations by providing a comprehensive characterization of lung‑relevant dark‑field signals, developing a flexible simulation environment to evaluate system configurations under realistic fabrication constraints, and constructing a first analyzer‑less prototype to simplify system design. To further improve manufacturability and enable compact geometries, diffraction beamlet arrays were implemented as an alternative grating approach.
In future clinical pathways, AI‑driven analysis and reconstruction methods may complement these technological developments by enhancing fringe extraction, assisting with automated disease detection, and supporting data‑driven optimization of imaging parameters—further strengthening the clinical value and usability of dark‑field lung imaging systems.
