自适应注意力引导的LDCT图像去噪条件扩散模型

doi:10.3778/j.issn.1002-8331.2406-0150

摘要/Abstract

摘要： 相较于生成对抗网络（GAN）和卷积神经网络（CNN），扩散模型在图像生成中表现出更高的图像质量以及更稳定的训练过程。然而，在去除低剂量计算机断层扫描（LDCT）图像噪声和伪影的任务中，其生成结果的多样性会导致LDCT图像细节失真。虽然分类器引导（classifier guidance，CG）和无分类器引导（classifier-free guidance，CFG）控制了其多样性，但引入了额外的训练需求和对外部条件的依赖。提出了自适应注意力引导的条件扩散模型AACD（adaptive attention-guided conditional diffusion model），在保证生成图像一致性的同时，减少了额外训练，并降低了对外部条件的依赖。为进一步缓解图像细节失真问题，还在U-Net的下采样层基础上设计了多尺度上下文感知网络MSCAN（multi-scale context-aware network）。MSCAN通过下采样处理，关注并融合多尺度信息，有效提升了图像的细节。实验结果表明，MSCAN和AACD在峰值信噪比（PSNR）、结构相似性（SSIM）和均方根误差（RMSE）这三个指标上展现出了较强的竞争力。

关键词: 低剂量CT, 医学图像去噪, 扩散模型, 引导采样

Abstract: Diffusion models have demonstrated higher image quality and more stable training process compared to generative adversarial networks (GANs) and convolutional neural networks (CNNs) in the task of image generation. However, the diversity of their generation results can lead to distortionss of details in low-dose computed tomography (LDCT) images. While classifier guidance (CG) and classifier-free guidance (CFG) control the diversity, they introduce additional training requirements and dependencies on external conditions. This study proposes an adaptive attention-guided conditional diffusion model (AACD) to ensure consistency of generated images with reduced extra training and lower dependence on external conditions. To further mitigate image detail distortion, this paper designs a multi-scale context-aware network (MSCAN) based on the downsampling layers of U-Net. MSCAN effectively enhances image details through downsampling processing, focusing on and fusing multi-scale information. Experimental results demonstrate that MSCAN and AACD exhibit strong competitiveness in terms of peak signal-to-noise ratio (PSNR), structural similarity (SSIM) and root mean square error (RMSE).

Key words: low-dose CT, medical image denoising, diffusion model, guided sampling

王少琦, 降爱莲, 马建芬. 自适应注意力引导的LDCT图像去噪条件扩散模型[J]. 计算机工程与应用, 2025, 61(19): 237-248.

WANG Shaoqi, JIANG Ailian, MA Jianfen. Adaptive Attention-Guided Conditional Diffusion Model for LDCT Image Denoising[J]. Computer Engineering and Applications, 2025, 61(19): 237-248.

参考文献

[1] 章云港, 易本顺, 吴晨玥, 等. 基于卷积神经网络的低剂量CT图像去噪方法[J]. 光学学报, 2018, 38(4): 123-129.
ZHANG Y G, YI B S, WU C Y, et al. Low-dose CT image denoising method based on convolutional neural network[J]. Acta Optica Sinica, 2018, 38(4): 123-129.
[2] CHEN H, ZHANG Y, KALRA M K, et al. Low-dose CT with a residual encoder-decoder convolutional neural network[J]. IEEE Transactions on Medical Imaging, 2017, 36(12): 2524-2535.
[3] YANG Q S, YAN P K, ZHANG Y B, et al. Low-dose CT image denoising using a generative adversarial network with Wasserstein distance and perceptual loss[J]. IEEE Transactions on Medical Imaging, 2018, 37(6): 1348-1357.
[4] HUANG Z Z, ZHANG J P, ZHANG Y, et al. DU-GAN: generative adversarial networks with dual-domain U-net-based discriminators for low-dose CT denoising[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 71: 4500512.
[5] HO J, JAIN A, ABBEEL P. Denoising diffusion probabilistic models[C]//Advances in Neural Information Processing Systems (NeurIPS), 2020: 6840-6851.
[6] DHARIWAL P, NICHOL A. Diffusion models beat GANs on image synthesis[C]//Advances in Neural Information Processing Systems (NeurIPS), 2021: 8780-8794.
[7] HO J, SAHARIA C, CHAN W, et al. Cascaded diffusion models for high fidelity image generation[J]. Journal of Machine Learning Research, 2022, 23(47): 1-33.
[8] ROMBACH R, BLATTMANN A, LORENZ D, et al. High-resolution image synthesis with latent diffusion models[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 10674-10685.
[9] NICHOL A, DHARIWAL P, RAMESH A, et al. GLIDE: towards photorealistic image generation and editing with text-guided diffusion models[J]. arXiv:2112.10741, 2021.
[10] HO J, SALIMANS T. Classifier-free diffusion guidance[C]//Proceedings of the NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications, 2021.
[11] ZHANG B W, GU S Y, ZHANG B, et al. StyleSwin: transformer-based GAN for high-resolution image generation[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 11294-11304.
[12] HOOGEBOOM E, SALIMANS T. Blurring diffusion models[J]. arXiv:2209.05557, 2022.
[13] NICHOL A Q, DHARIWAL P. Improved denoising diffusion probabilistic models[C]//Proceedings of the International Conference on Machine Learning, 2021: 8162-8171.
[14] SONG J, MENG C, ERMON S. Denoising diffusion implicit models[C]//Proceedings of the International Conference on Learning Representations, 2021.
[15] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems (NeurIPS), 2017.
[16] ZHANG H, GOODFELLOW I, METAXAS D, et al. Self-attention generative adversarial networks[C]//Proceedings of the International Conference on Machine Learning, 2019: 7354-7363.
[17] JIANG Y, CHANG S, WANG Z. TransGAN: two pure transformers can make one strong GAN, and that can scale up[C]//Advances in Neural Information Processing Systems (NeurIPS), 2021: 14745-14758.
[18] PEEBLES W, XIE S. Scalable diffusion models with transformers[J]. arXiv:2212.09748, 2022.
[19] CHEN L C, PAPANDREOU G, KOKKINOS I, et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848.
[20] HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 7132-7141.
[21] WANG Q L, WU B G, ZHU P F, et al. ECA-net: efficient channel attention for deep convolutional neural networks[C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2020: 11531-11539.
[22] LIU W, RABINOVICH A, BERG A C. ParseNet: looking wider to see better[J]. arXiv:1506.04579, 2015.
[23] DAI Y M, GIESEKE F, OEHMCKE S, et al. Attentional feature fusion[C]//Proceedings of the 2021 IEEE Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2021: 3559-3568.
[24] CHEN B Y, LENG S, YU L F, et al. An open library of CT patient projection data[C]//Proceedings of SPIE, 2016: 330-335.
[25] WANG D Y, FAN F L, WU Z, et al. CTformer: convolution-free Token2Token dilated vision transformer for low-dose CT denoising[J]. Physics in Medicine and Biology, 2023, 68(6): 065012.
[26] LIU X, XIE Y, CHENG J, et al. Diffusion probabilistic priors for zero-shot low-dose CT image denoising[J]. arXiv:2305.
15887, 2023.