Filter2Noise: A Framework for Interpretable and Zero-Shot Low-Dose CT Image Denoising

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Sun, Yipeng, Schneider, Linda-Sophie, Mei, Siyuan, Wang, Jinhua, Hu, Ge, Gu, Mingxuan, Ye, Chengze, Wagner, Fabian, Song, Lan, Bayer, Siming, Maier, Andreas
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918343405993984
author Sun, Yipeng
Schneider, Linda-Sophie
Mei, Siyuan
Wang, Jinhua
Hu, Ge
Gu, Mingxuan
Ye, Chengze
Wagner, Fabian
Song, Lan
Bayer, Siming
Maier, Andreas
author_facet Sun, Yipeng
Schneider, Linda-Sophie
Mei, Siyuan
Wang, Jinhua
Hu, Ge
Gu, Mingxuan
Ye, Chengze
Wagner, Fabian
Song, Lan
Bayer, Siming
Maier, Andreas
contents Noise in low-dose computed tomography (LDCT) can obscure important diagnostic details. While deep learning offers powerful denoising, supervised methods require impractical paired data, and self-supervised alternatives often use opaque, parameter-heavy networks that limit clinical trust. We propose Filter2Noise (F2N), a novel self-supervised framework for interpretable, zero-shot denoising from a single LDCT image. Instead of a black-box network, its core is an Attention-Guided Bilateral Filter, a transparent, content-aware mathematical operator. A lightweight attention module predicts spatially varying filter parameters, making the process transparent and allowing interactive radiologist control. To learn from a single image with correlated noise, we introduce a multi-scale self-supervised loss coupled with Euclidean Local Shuffle (ELS) to disrupt noise patterns while preserving anatomical integrity. On the Mayo Clinic LDCT Challenge, F2N achieves state-of-the-art results, outperforming competing zero-shot methods by up to 3.68 dB in PSNR. It accomplishes this with only 3.6k parameters, orders of magnitude fewer than competing models, which accelerates inference and simplifies deployment. By combining high performance with transparency, user control, and high parameter efficiency, F2N offers a trustworthy solution for LDCT enhancement. We further demonstrate its applicability by validating it on clinical photon-counting CT data. Code is available at: https://github.com/sypsyp97/Filter2Noise.
format Preprint
id arxiv_https___arxiv_org_abs_2504_13519
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Filter2Noise: A Framework for Interpretable and Zero-Shot Low-Dose CT Image Denoising
Sun, Yipeng
Schneider, Linda-Sophie
Mei, Siyuan
Wang, Jinhua
Hu, Ge
Gu, Mingxuan
Ye, Chengze
Wagner, Fabian
Song, Lan
Bayer, Siming
Maier, Andreas
Image and Video Processing
Computer Vision and Pattern Recognition
Machine Learning
Noise in low-dose computed tomography (LDCT) can obscure important diagnostic details. While deep learning offers powerful denoising, supervised methods require impractical paired data, and self-supervised alternatives often use opaque, parameter-heavy networks that limit clinical trust. We propose Filter2Noise (F2N), a novel self-supervised framework for interpretable, zero-shot denoising from a single LDCT image. Instead of a black-box network, its core is an Attention-Guided Bilateral Filter, a transparent, content-aware mathematical operator. A lightweight attention module predicts spatially varying filter parameters, making the process transparent and allowing interactive radiologist control. To learn from a single image with correlated noise, we introduce a multi-scale self-supervised loss coupled with Euclidean Local Shuffle (ELS) to disrupt noise patterns while preserving anatomical integrity. On the Mayo Clinic LDCT Challenge, F2N achieves state-of-the-art results, outperforming competing zero-shot methods by up to 3.68 dB in PSNR. It accomplishes this with only 3.6k parameters, orders of magnitude fewer than competing models, which accelerates inference and simplifies deployment. By combining high performance with transparency, user control, and high parameter efficiency, F2N offers a trustworthy solution for LDCT enhancement. We further demonstrate its applicability by validating it on clinical photon-counting CT data. Code is available at: https://github.com/sypsyp97/Filter2Noise.
title Filter2Noise: A Framework for Interpretable and Zero-Shot Low-Dose CT Image Denoising
topic Image and Video Processing
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2504.13519