Abstract

To propose a CT truncated data reconstruction model (DDTrans) based on projection and image dualdomain Transformer coupled feature learning for reducing truncation artifacts and image structure distortion caused by insufficient field of view (FOV) in CT scanning.