Anatomy-Aware Low-Dose CT Denoising via Pretrained Vision Models and Semantic-Guided Contrastive Learning

2025-08-11

AI Summary

Existing low-dose CT (LDCT) denoising methods often neglect anatomical semantics, leading to structural distortion and over-smoothing. To address this, we propose ALDEN—a generative adversarial denoising framework that integrates semantic priors from pre-trained vision models. Its core innovations include: (1) an anatomy-aware discriminator employing cross-attention to fuse multi-level semantic features; and (2) a semantic-guided contrastive learning module that enhances tissue consistency and suppresses artifacts using positive/negative sample pairs. ALDEN is the first method to achieve tissue-specific noise suppression driven by hierarchical anatomical information. Evaluated on two public LDCT datasets, it achieves state-of-the-art quantitative performance while significantly mitigating over-smoothing. Furthermore, downstream organ segmentation across 117 anatomical classes demonstrates substantial improvement in anatomical structure preservation, validating its clinical relevance and fidelity.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Generation

Application Category

• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

To reduce radiation exposure and improve the diagnostic efficacy of low-dose computed tomography (LDCT), numerous deep learning-based denoising methods have been developed to mitigate noise and artifacts. However, most of these approaches ignore the anatomical semantics of human tissues, which may potentially result in suboptimal denoising outcomes. To address this problem, we propose ALDEN, an anatomy-aware LDCT denoising method that integrates semantic features of pretrained vision models (PVMs) with adversarial and contrastive learning. Specifically, we introduce an anatomy-aware discriminator that dynamically fuses hierarchical semantic features from reference normal-dose CT (NDCT) via cross-attention mechanisms, enabling tissue-specific realism evaluation in the discriminator. In addition, we propose a semantic-guided contrastive learning module that enforces anatomical consistency by contrasting PVM-derived features from LDCT, denoised CT and NDCT, preserving tissue-specific patterns through positive pairs and suppressing artifacts via dual negative pairs. Extensive experiments conducted on two LDCT denoising datasets reveal that ALDEN achieves the state-of-the-art performance, offering superior anatomy preservation and substantially reducing over-smoothing issue of previous work. Further validation on a downstream multi-organ segmentation task (encompassing 117 anatomical structures) affirms the model's ability to maintain anatomical awareness.

Problem

• Reducing noise and artifacts in low-dose CT scans
• Preserving anatomical semantics during CT denoising
• Preventing over-smoothing while maintaining tissue-specific patterns

Innovation

• Integrates pretrained vision models with adversarial learning
• Uses anatomy-aware discriminator with cross-attention mechanisms
• Implements semantic-guided contrastive learning for anatomical consistency