Unsupervised low-dose CT reconstruction with one-way conditional normalizing flows
An, Ran and Li, Lemeng and Chen, Ke and Li, Hongwei (2025) Unsupervised low-dose CT reconstruction with one-way conditional normalizing flows. IEEE Transactions on Computational Imaging. ISSN 2333-9403 (In Press)
![]() |
Text.
Filename: An-etal-IEEE-TCI-2025-Unsupervised-Low-dose-CT-Reconstruction-with-one-way.pdf
Accepted Author Manuscript Restricted to Repository staff only until 1 January 2099. Download (5MB) | Request a copy |
Abstract
Deep-learning methods have shown promising performance in low-dose computed tomography (LDCT) reconstruction. However, supervised methods face the problem of lacking labeled data in actual scenarios, and the CNN-based unsupervised denoising methods would cause excessive smoothing in the reconstructed image. Recently, the normalizing flows (NFs) based methods have shown advantages in producing detail-rich images and avoiding over-smoothing, however, there are still issues: (1) Although the alternating optimization in the data and latent space can well utilize the regularization and generation capabilities of NFs, the current two-way transformation strategy of noisy images and latent variables would cause secondary artifacts; and (2) Training NFs on high-resolution CT images is hard due to huge computation, although using conditional normalizing flows (CNFs) to learn conditional probability can reduce the computational burden, current methods require labeled data for conditionalization, and the unsupervised CNFsbased LDCT reconstruction remains a problem. To tackle these problems, we propose a novel CNFs-based unsupervised LDCT iterative reconstruction algorithm. It employs strict one-way transformation when performing alternating optimization in the dual spaces, thus effectively avoiding the secondary artifacts. By proposing a novel unsupervised conditionalization strategy for LDCT images, we train CNFs on high-resolution images thus achieving fast and high-quality unsupervised reconstruction. Experiments on different datasets suggest that the performance of the proposed algorithm could surpass some state-of-the-art unsupervised and even supervised methods.
ORCID iDs
An, Ran, Li, Lemeng, Chen, Ke
-
-
Item type: Article ID code: 92302 Dates: DateEvent9 March 2025Published9 March 2025AcceptedSubjects: Science > Mathematics > Electronic computers. Computer science Department: Faculty of Science > Mathematics and Statistics Depositing user: Pure Administrator Date deposited: 11 Mar 2025 10:38 Last modified: 11 Mar 2025 10:38 URI: https://strathprints.strath.ac.uk/id/eprint/92302