FDDM: Unsupervised Medical Image Translation with a Frequency-Decoupled Diffusion Model

arXiv:2311.12070v2 Announce Type: replace-cross
Abstract: Diffusion models have demonstrated significant potential in producing high-quality images in medical image translation to aid disease diagnosis, localization, and treatment. Nevertheless, current diffusion models have limited success in achieving faithful image translations that can accurately preserve the anatomical structures of medical images, especially for unpaired datasets. The preservation of structural and anatomical details is essential to reliable medical diagnosis and treatment planning, as structural mismatches can lead to disease misidentification and treatment errors. In …

abstract arxiv cs.cv current diagnosis diffusion diffusion model diffusion models disease disease diagnosis eess.iv image images localization medical potential quality replace success translation translations treatment type unsupervised