Zero-Shot Low-dose CT Denoising via Sinogram Flicking

2025-04-10

AI Summary

Low-dose CT denoising faces dual challenges: the absence of paired ground-truth labels and the resolution degradation inherent in single-image self-supervised methods. To address these, we propose the first sinogram-domain zero-shot self-supervised method. Leveraging the conjugate-ray noise statistical properties, we design a sinogram flicking mechanism that generates pseudo-paired sinograms—content-consistent yet noise-diverse—via random block swapping. We further introduce a lightweight ZS-NSN network enabling end-to-end training from a single noisy sinogram. Crucially, our approach avoids spatial downsampling, preserving full native resolution, and requires neither clinically infeasible clean data nor auxiliary images. Extensive simulations demonstrate significant improvements in PSNR and SSIM over state-of-the-art zero-shot methods (e.g., ZS-N2N), with superior high-frequency detail reconstruction. Our method effectively overcomes key bottlenecks in single-image self-supervision—namely, data diversity limitation and spatial fidelity degradation.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: Other Foundations of Computer Vision
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Many low-dose CT imaging methods rely on supervised learning, which requires a large number of paired noisy and clean images. However, obtaining paired images in clinical practice is challenging. To address this issue, zero-shot self-supervised methods train denoising networks using only the information within a single image, such as ZS-N2N. However, these methods often employ downsampling operations that degrade image resolution. Additionally, the training dataset is inherently constrained to the image itself. In this paper, we propose a zero-shot low-dose CT imaging method based on sinogram flicking, which operates within a single image but generates many copies via random conjugate ray matching. Specifically, two conjugate X-ray pencil beams measure the same path; their expected values should be identical, while their noise levels vary during measurements. By randomly swapping portions of the conjugate X-rays in the sinogram domain, we generate a large set of sinograms with consistent content but varying noise patterns. When displayed dynamically, these sinograms exhibit a flickering effect due to their identical structural content but differing noise patterns-hence the term sinogram flicking. We train the network on pairs of sinograms with the same content but different noise distributions using a lightweight model adapted from ZS-NSN. This process is repeated to obtain the final results. A simulation study demonstrates that our method outperforms state-of-the-art approaches such as ZS-N2N.

Problem

• Eliminates need for paired images in low-dose CT denoising
• Avoids resolution loss from downsampling in zero-shot methods
• Generates diverse noise patterns via sinogram flicking technique

Innovation

• Zero-shot denoising via sinogram flicking
• Random conjugate ray matching for noise variation
• Lightweight model training on noise-varying sinograms