Filter2Noise: Interpretable Self-Supervised Single-Image Denoising for Low-Dose CT with Attention-Guided Bilateral Filtering

2025-04-18

AI Summary

Low-dose CT single-image denoising faces challenges including scarcity of supervised training data, reliance of self-supervised methods on multiple noisy realizations, and poor model interpretability. Method: We propose an interpretable self-supervised framework featuring: (i) an attention-guided bilateral filter—novel in enabling post-hoc visualization and region-level parameter adjustment; (ii) a single-image Noise2Noise extension incorporating downsampling-based shuffling to model spatially correlated noise; and (iii) a lightweight spatial parameter prediction module coupled with a self-supervised reconstruction loss. Results: On the Mayo Clinic 2016 dataset, our method achieves a 4.59 dB PSNR gain over ZS-N2N. It delivers high denoising accuracy, strong interpretability, user-controllable parameter tuning, and parameter efficiency—establishing a new paradigm for precise clinical diagnosis of subtle anatomical structures and low-contrast lesions.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Machine Learning: Unsupervised & Self-Supervised Learning
• Natural Language Processing: Safety and Robustness

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Effective denoising is crucial in low-dose CT to enhance subtle structures and low-contrast lesions while preventing diagnostic errors. Supervised methods struggle with limited paired datasets, and self-supervised approaches often require multiple noisy images and rely on deep networks like U-Net, offering little insight into the denoising mechanism. To address these challenges, we propose an interpretable self-supervised single-image denoising framework -- Filter2Noise (F2N). Our approach introduces an Attention-Guided Bilateral Filter that adapted to each noisy input through a lightweight module that predicts spatially varying filter parameters, which can be visualized and adjusted post-training for user-controlled denoising in specific regions of interest. To enable single-image training, we introduce a novel downsampling shuffle strategy with a new self-supervised loss function that extends the concept of Noise2Noise to a single image and addresses spatially correlated noise. On the Mayo Clinic 2016 low-dose CT dataset, F2N outperforms the leading self-supervised single-image method (ZS-N2N) by 4.59 dB PSNR while improving transparency, user control, and parametric efficiency. These features provide key advantages for medical applications that require precise and interpretable noise reduction. Our code is demonstrated at https://github.com/sypsyp97/Filter2Noise.git .

Problem

• Develop interpretable self-supervised denoising for low-dose CT
• Overcome limited paired datasets with single-image training
• Enhance user control and transparency in noise reduction

Innovation

• Attention-Guided Bilateral Filter for interpretable denoising
• Lightweight module predicts spatially varying filter parameters
• Novel downsampling shuffle strategy for single-image training