Development of an AI-Based Algorithm for the Correction of Moiré Artifacts in Digital Radiography – MT Intro Talk by Shadi Khamseh

Join us for an introductory presentation of a Master’s thesis on an AI-based algorithm for correcting Moiré artifacts in digital radiography. The talk will begin by outlining how anti-scatter grids and detector configurations can generate high-frequency Moiré patterns that overlap diagnostically relevant structures and cannot be effectively removed by standard filtering techniques. The study will assemble a dataset of radiographic images spanning multiple detector types and anatomical regions. A deep learning artifact-correction algorithm will be developed to identify and suppress these patterns while preserving image fidelity. Finally, evaluation metrics and experimental protocols – including variations in noise levels, grid configurations, and reconstruction parameters – will be defined to assess performance and support potential integration into clinical imaging workflows.