LDCT Denoising
LDCT: Low-Dose Computed Tomography
Overview
LDCT denoising studies how to suppress noise caused by reduced radiation exposure without obscuring anatomy or lesions. Approaches include CNNs and transformers, diffusion and autoregressive models, and physics-informed methods that learn from projection data; self-supervised strategies aim to work when paired clean scans are unavailable. A central challenge is generalizing across dose levels, scanners, and anatomical regions, while balancing efficient inference with realistic detail recovery. Standard metrics such as PSNR and SSIM can reward smoother images even when lesion detectability suffers, so clinical and task-based evaluation remains an important open issue.
Latest papers
24 papers3 papers in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.