Efficient Noise Context-Aware Representation for Low-Dose CT Denoising
Abstract
While deep learning-based denoising has become widely adopted in low-dose CT, conventional models use generic architectures designed for natural images, failing to account for non-stationary and spatially correlated CT noise characteristics. To address this, we propose an Efficient Noise COntext-aware REpresentation (ENCORE) framework that explicitly leverages CT noise characteristics and anatomical features. First, we reformulate the noise synthesis procedure based on a realistic noise distribution beyond the conventional Gaussian approximation, establishing a rigorous foundation for training pair generation. Next, we extract local noise power and correlation contexts to guide the denoising process. To fully leverage the potential of noise context, we propose a FlyingConv module, which adaptively changes convolution weights for each local image region. Notably, our approach demonstrates substantial gains in both denoising quality and computational efficiency. Furthermore, manipulating the intensity of the noise context maps at inference time enables zero-shot conditional denoising, allowing for dynamic control over the output image texture. The entire pipeline is available at https://github.com/minwoo-yu/ENCORE.git
Related
- Hybrid neural denoising for resource-efficient near- and sub-threshold radio triggering of extensive air showers
- Acoustic coherence-aware oriented 3D photoacoustic microscopy super-resolution with noise-suppression
- Noise-aware emulation and cross-device validation of neutral atom analog quantum processing units
- Efficient angle-aware differentiable Mask3D modeling for inverse lithography
- Representation-Aware Transport-Information Measure for Non-inclusive Discrete Supports
BibTeX
@misc{yu2026efficient,
title = {{Efficient Noise Context-Aware Representation for Low-Dose CT Denoising}},
author = {Minwoo Yu and N. Robert Bennett and Jongduk Baek and Adam S. Wang},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2608.10343},
eprint = {2608.10343},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.10343},
}
RIS
TY - UNPB
TI - Efficient Noise Context-Aware Representation for Low-Dose CT Denoising
AU - Minwoo Yu
AU - N. Robert Bennett
AU - Jongduk Baek
AU - Adam S. Wang
PY - 2026
DA - 2026/10/14
PB - arXiv
DO - 10.48550/arxiv.2608.10343
UR - https://arxiv.org/abs/2608.10343
AB - While deep learning-based denoising has become widely adopted in low-dose CT, conventional models use generic architectures designed for natural images, failing to account for non-stationary and spatially correlated CT noise characteristics. To address this, we propose an Efficient Noise COntext-aware REpresentation (ENCORE) framework that explicitly leverages CT noise characteristics and anatomical features. First, we reformulate the noise synthesis procedure based on a realistic noise distribution beyond the conventional Gaussian approximation, establishing a rigorous foundation for training pair generation. Next, we extract local noise power and correlation contexts to guide the denoising process. To fully leverage the potential of noise context, we propose a FlyingConv module, which adaptively changes convolution weights for each local image region. Notably, our approach demonstrates substantial gains in both denoising quality and computational efficiency. Furthermore, manipulating the intensity of the noise context maps at inference time enables zero-shot conditional denoising, allowing for dynamic control over the output image texture. The entire pipeline is available at https://github.com/minwoo-yu/ENCORE.git
ER -