基于优化生成对抗网络的OCT图像去噪和超分辨率重建

doi:10.12066/j.issn.1007-2861.2566

摘要/Abstract

摘要： 光学相干层析成像(optical coherence tomography,OCT)运用低相干光源,得到的图像会不可避免地受到散斑噪声的影响.为了获得高信噪比和高分辨率的OCT图像,提出了一种基于生成对抗网络的超分辨率重建网络模型,以同时实现对OCT图像的去噪和超分辨率重建.将所提模型在OCT图像数据集上进行实验,并与一些著名的模型进行了定量和定性比较.研究结果表明,该模型峰值信噪比(peak signal-to-noise ratio,PSNR)平均值表现居中,学习感知图像块相似度平均值表现优越,表明该模型能更好地恢复图像细节,更有利于进行医学图像诊断,提高临床诊断准确率.

关键词: 光学相干层析成像, 超分辨率, 图像去噪, 生成对抗网络

Abstract: Optical coherence tomography (OCT) uses a low-coherence optical source, and the images obtained are inevitably affected by scattering noise. To obtain OCT images with high signal-to-noise ratio and high resolution, a super-resolution reconstruction network model is proposed based on the generative adversarial network to simultaneously achieve denoising and super-resolution reconstruction of OCT images. This model is evaluated on an OCT image dataset and compared with some well-established models quantitatively and qualitatively. The results show that the average peak signal-to-noise ratio (PSNR) of this model is in the intermediate range and that the average similarity value of the learned perceptual image block is superior. This indicates that this model effectively recovers the image details and is conducive to the diagnosis of medical images, thus improving the accuracy of clinical diagnosis.

Key words: optical coherence tomography (OCT), super-resolution, image denoising, generative adversarial network

中图分类号:

TP391

赵静, 王驰, 俞朱恺, 许婧靓. 基于优化生成对抗网络的OCT图像去噪和超分辨率重建[J]. 上海大学学报(自然科学版), 2026, 32(2): 251-260.

ZHAO Jing, WANG Chi, YU Zhukai, XU Jingjing. OCT image denoising and super-resolution reconstruction based on optimized generative adversarial networks[J]. Journal of Shanghai University（Natural Science Edition）, 2026, 32(2): 251-260.

参考文献

[1] Huang D, Swanson E A, Lin C P, et al. Optical coherence tomography [J]. Science, 1991, 254(5053): 1178-1181.
[2] 王海霞, 杨熙丞, 梁荣华, 等. 基于轻量级U-Net神经网络提取OCT内指纹和外指纹[J]. 高技术通讯, 2019, 29(11): 1063-1072.
[3] Zeng N, He Y H, Ma H, et al. Pearl detection with optical coherence tomography [J]. Chinese Journal of Lasers, 2007, 34(8): 1140-1145.
[4] Motschi A R, Roberts P K, Desissaire S, et al. Identification and quantification of fibrotic areas in the human retina using polarization-sensitive OCT [J]. Biomedical Optics Express, 2021, 12(7): 4380-4400.
[5] Wang J, Xu Y, Mesa K J, et al. Complementary use of polarization-sensitive and standard OCT metrics for enhanced intraoperative differentiation of breast cancer [J]. Biomedical Optics Express, 2018, 9(12): 6519-6528.
[6] Baumann B. Polarization sensitive optical coherence tomography: a review of technology and applications [J]. Applied Sciences, 2017, 7(5): 474-507.
[7] Li Y, Chen J S, Chen Z P. Advances in doppler optical coherence tomography and angiography [J]. Translational Biophotonics, 2019. DOI: 10.1002/tbio.201900005.
[8] Schmitt J M, Xiang S, Yung K M. Speckle in optical coherence tomography [J]. Journal of Biomedical Optics, 1999, 4(1): 95-105.
[9] Yang Q X, Yang R G, Davis J, et al. Spatial-depth super resolution for range images [C]// IEEE Conference on Computer Vision & Pattern Recognition. 2007: 1-8.
[10] Chen H, Zhang Y, Zhang W H, et al. Low-dose ct via convolutional neural network [J]. Biomedical Optics Express, 2017, 8(2): 679-694.
[11] Nasrollahi K, Moeslund B T. Super-resolution: a comprehensive survey [J]. Mach Vis Appl, 2014, 25(6): 1423-1468.
[12] Dong C, Loy C C, He K, et al. Image super-resolution using deep convolutional networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, 38(2): 295-307.
[13] Sajjadi M S M, Scholkopf B, Hirsch M. EnhanceNet: single image super-resolution through automated texture synthesis [C]// Proceedings of the IEEE International Conference on Computer Vision (ICCV). 2017: 4491-4500.
[14] Kim J, Lee J K, Lee K M. Accurate image super-resolution using very deep convolutional networks [C]// IEEE Conferenceon Computer Vision & Pattern Recognition. 2016: 1646-1654.
[15] Kim J, Lee J K, Lee K M. Deeply-recursive convolutional network for image superresolution [C]// IEEE Conference on Computer Vision & Pattern Recognition. 2016: 1637-1645.
[16] Lai W S, Huang J B, Ahuja N, et al. Deep laplacian pyramid networks for fast and accurate super-resolution [C]// IEEE Conference on Computer Vision & Pattern Recognition. 2017: 5835-5843.
[17] Wang X T, Yu K, Wu S X, et al. ESRGAN: enhanced super-resolution generative adversarial networks [C]// European Conference on Computer Vision Workshops. 2018: 63-79.
[18] Ledig C, Theis L, Huszar F, et al. Photo-realistic single image super-resolution using a generative adversarial network [C]// IEEE Conference on Computer Vision & Pattern Recognition. 2017: 105-114.
[19] Huang Y Q, Lu Z X, Shao Z M, et al. Simultaneous denoising and super-resolution of optical coherence tomography images based on generative adversarial network [J]. Optics Express, 2019, 27(9): 12289-12307.
[20] Rakotonirina N C, Rasoanaivo A. ESRGAN+: further improving enhanced superresolution generative adversarial network [C]// IEEE International Conference on Acoustics, Speech and Signal Processing. 2020: 3637-3641.
[21] Haris M, Shakhnarovich G, Ukita N. Deep back-projection networks for superresolution [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 1664-1673.
[22] Jolicoeur-Martineau A. The relativistic discriminator: a key element missing from standard GAN [C]// International Conference on Learning Representations. 2019.
[23] Justin J, Alexandre A, Li F F. Perceptual losses for real-time style transfer and superresolution [C]// European Conference on Computer Vision. 2016: 694-711.
[24] Krizhevsky A, Sutskever I, Hinton G E. Imagenet classification with deep convolutional neural networks [C]// Advances in Neural Information Processing Systems. 2012: 1097-1105.
[25] Bruna J, Sprechmann P, Lecun Y. Super-resolution with deep convolutional sufficient statistics [J]. Computer Science, 2015. DOI: 10.48550/arXiv.1511.05666.
[26] Fang L, Li S, McNabb R P, et al. Fast acquisition and reconstruction of optical coherence tomography images via sparse representation [J]. IEEE Transactions on Medical Imaging, 2013, 32(11): 2034-2049.
[27] Sina F, Stephanie J C, Rachelle V O, et al. Quantitative classification of eyes with and without intermediate age-related macular degeneration using optical coherence tomography [J]. Ophthalmology, 2014, 121(1): 162-172.
[28] Wang Z X, Gao G W, Li J C, et al. Lightweight image super-resolution with multi-scale feature interaction network [C]// IEEE International Conference on Multimedia and Expo. 2021: 1-6.
[29] Yang C Y, Ma C, Yang M H. Single-image super-resolution: a benchmark [C]// European Conference on Computer Vision. 2014: 372-386.
[30] Zhang R, Isola P, Efros A A, et al. The unreasonable effectiveness of deep features as a perceptual metric [C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2018: 586-595.
[31] Wang Z, Simoncelli E P, Bovik A C. Multi-scale structural similarity for image quality assessment [C]// IEEE Asilomar Conference on Signals, Systems and Computers. 2003: 9-13.
[32] Wang Z, Bovik A C, Sheikh H R, et al. Image quality assessment: from error visibility to structural similarity [J]. IEEE Transactions on Image Processing, 2004, 13(4): 600-612.
[33] Gupta P, Srivastava P, Bhardwaj S, et al. A modified PSNR metric based on HVS for quality assessment of color images [C]// IEEE International Conference on Communication and Industrial Application. 2011: 96-99.