基于卷积神经网络的测序图像超分辨重建

高通量测序在过去的十多年里迅速发展并被广泛应用，主流的高通量测序技术运用了“边合成边测序”的思想，在合成的同时让碱基本身产生荧光。基于超分辨显微成像技术的测序仪可以使得DNA样品在载体上排布更加紧密，提高测序通量并降低测序成本。但传统的结构光照明显微重建算法存在重建伪影、信噪比低、重建速度慢等问题，不满足基因测序对准确性和速度的要求。

本文基于经典超分辨网络RCAN对基因测序图像进行超分辨重建，使用模拟的测序图像对深度学习模型进行训练，并对实验中拍得的低分辨率测序图片进行重建，验证了模拟训练集的可行性和网络的泛化性，避免了通过实验拍照获得训练数据的繁琐步骤。由于测序通量对图像重建速度的要求，本论文基于RCAN网络进行模型结构的简化，在保证重建精度的前提下，验证了简化后网络良好的重建效果，将重建速度提升了1000多倍。本文还在RCAN简化网络的基础上，提出一个基于重参数残差通道注意力机制的轻量级超分辨模型L-Rercan，在重建速度相同的情况下，运用信息通道整合模块对通道之间的信息进行线性组合，提高了模型的表达能力，获得了更好的图像重建效果。

[1] BIRNEY E. The international human genome project[J]. Human Molecular Genetics, 2021, 30(R2): R161-R163.
[2] STARK R, GRZELAK M, HADFIELD J. RNA sequencing: the teenage years[J]. Nature Reviews Genetics, 2019, 20(11): 631-656.
[3] XUE R, LI R, BAI F. Single cell sequencing: technique, application, and future development[J]. Science Bulletin, 2015, 60(1): 33-42.
[4] BHARTI R, GRIMM D G. Current challenges and best-practice protocols for microbiome analysis[J]. Briefings in Bioinformatics, 2021, 22(1): 178-193.
[5] SANGER F, AIR G M, BARRELL B G, et al. Nucleotide sequence of bacteriophage φX174 DNA[J]. Nature (London), 1977, 265(5596): 687-695.
[6] NELLIMARLA S, KESANAKURTI P. Next-generation sequencing: a promising tool for vaccines and other biological products[J]. Vaccines, 2023, 11(3): 527.
[7] MCCARTHY A. Third generation DNA sequencing: pacific biosciences' single molecule real time technology[J]. Chemistry & Biology, 2010, 17(7): 675-676.
[8] BRANTON D, DEAMER D W, MARZIALI A, et al. The potential and challenges of nanopore sequencing[J]. Nature Biotechnology, 2008, 26(10): 1146-1153.
[9] RAO J, PENG L, LIANG X, et al. Performance of copy number variants detection based on whole-genome sequencing by DNBSEQ platforms[J]. BMC Bioinformatics, 2020, 21(1): 1-14.
[10] ABBE E. Beiträge zur Theorie des mikroskops und der mikroskopischen wahrnehmung[J]. Archiv Für Mikroskopische Anatomie, 1873, 9(1): 413-468.
[11] GUSTAFSSON M G L, AGARD D A, SEDAT J W. Doubling the lateral resolution of wide-field fluorescence microscopy using structured illumination[C]//Three-Dimensional and Multidimensional Microscopy: Image Acquisition Processing VII. SPIE, 2000, 3919: 141-150.
[12] HELL S W, WICHMANN J. Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion fluorescence microscopy[J]. Optics Letters, 1994, 19(11): 780-782.
[13] ZHUANG X, RUST M J, BATES M. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM)[J]. Nature Methods, 2006, 3(10): 793-796.
[14] HESS S T, GIRIRAJAN T P K, MASON M D. Ultra-high resolution imaging by fluorescence photoactivation localization microscopy[J]. Biophysical Journal, 2006, 91(11): 4258-4272.
[15] GUSTAFSSON M G L, KNER P, CHHUN B B, et al. Super-resolution video microscopy of live cells by structured illumination[J]. Nature Methods, 2009, 6(5): 339-342.
[16] INGERMAN E A, LONDON R A, HEINTZMANN R, et al. Signal, noise and resolution in linear and nonlinear structured-illumination microscopy[J]. Journal of Microscopy (Oxford), 2019, 273(1): 3-25.
[17] FROHN J T, KNAPP H F, STEMMER A. True optical resolution beyond the rayleigh limit achieved by standing wave illumination[J]. Proceedings of the National Academy of Sciences, 2000, 97(13): 7232-7236.
[18] MUDRY E, BELKEBIR K, GIRARD J, et al. Structured illumination microscopy using unknown speckle patterns[J]. Nature Photonics, 2012, 6(5): 312-315.
[19] LAL A, SHAN C, XI P. Structured illumination microscopy image reconstruction Algorithm[J]. IEEE Journal of Selected Topics in Quantum Electronics, 2016, 22(4): 50-63.
[20] TU S, LIU Q, LIU X, et al. Fast reconstruction algorithm for structured illumination microscopy[J]. Optics Letters, 2020, 45(6): 1567-1570.
[21] DAN D, WANG Z, ZHOU X, et al. Rapid image reconstruction of structured illumination microscopy directly in the spatial domain[J]. IEEE Photonics Journal, 2021, 13(1): 1-11.
[22] WANG Z, TIANYU Z, HAO H, et al. High-speed image reconstruction for optically sectioned, super-resolution structured illumination microscopy[J]. Advanced Photonics, 2022, 4(2): 026003-026003.
[23] DEMMERLE J, INNOCENT C, NORTH A J, et al. Strategic and practical guidelines for successful structured illumination microscopy[J]. Nature Protocols, 2017, 12(5): 988-1010.
[24] ISOGAWA K, IDA T, SHIODERA T, et al. Deep shrinkage convolutional neural network for adaptive noise reduction[J]. IEEE Signal Processing Letters, 2018, 25(2): 224-228.
[25] ZHANG K, ZUO W, ZHANG L. FFDNet: toward a fast and flexible solution for cnn-based image denoising[J]. IEEE Transactions on Image Processing, 2018, 27(9): 4608-4622.
[26] CHEN J, CHEN J, CHAO H, et al. Image blind denoising with generative adversarial network based noise modeling[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 3155-3164.
[27] LI K, YANG S, DONG R, et al. Survey of single image super-resolution reconstruction[J]. IET Image Processing, 2020, 14(11): 2273-2290.
[28] DONG C, LOY C C, HE K, et al. Learning a deep convolutional network for image super-resolution[C]//Proceedings of European Conference on Computer Vision (ECCV), 2014: 184-199.
[29] LEDIG C, THEIS L, HUSZAR F, et al. Photo-realistic single image super-resolution using a generative adversarial network[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 4681-4690.
[30] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 770-778..
[31] LIM B, SON S, KIM H, et al. Enhanced deep residual networks for single image super-resolution[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2017: 136-144.
[32] ZHANG Y, LI K, LI K, et al. Image super-resolution using very deep residual channel attention networks[C]//Proceedings of the European Conference on Computer Vision (ECCV), 2018: 286-301.
[33] 孙叶美. 基于卷积神经网络的图像超分辨率重建算法研究[D]. 天津工业大学, 2019.
[34] 林青宇. 基于深度学习的单幅图像超分辨重建方法研究[D]. 北京建筑大学, 2022.
[35] 龙法宁, 朱晓姝, 胡春娇. 基于深层卷积网络的单幅图像超分辨率重建模型[J]. 广西科学, 2017, 24(3): 231-235.
[36] 丁文倩. 基于深度学习的单幅图像超分辨重建算法研究[D]. 东南大学, 2019.
[37] GUO M, LI Y, SU Y, et al. Rapid image deconvolution and multiview fusion for optical microscopy[J]. Nature Biotechnology, 2020, 38(11): 1337-1346.
[38] LUCY L B. An iterative technique for the rectification of observed distributions[J]. The Astronomical Journal, 1974, 79: 745-754.
[39] RICHARDSON W H. Bayesian-based iterative method of image restoration[J]. Journal of the Optical Society of America, 1972, 62(1): 55-59.
[40] JIN L, LIU B, ZHAO F, et al. Deep learning enables structured illumination microscopy with low light levels and enhanced speed[J]. Nature Communications, 2020, 11(1): 1934-1934.
[41] CHRISTENSEN C N, WARD E N, LIO P, et al. ML-SIM: universal reconstruction of structured illumination microscopy images using transfer learning[J]. Biomedical Optics Express, 2021, 12(5): 2720-2733.
[42] MüLLER M, MöNKEMöLLER V, HENNIG S, et al. Open-source image reconstruction of super-resolution structured illumination microscopy data in ImageJ[J]. Nature Communications, 2016, 7(1): 10980-10980.
[43] SHAH Z H, MüLLER M, WANG T C, et al. Deep-learning based denoising and reconstruction of super-resolution structured illumination microscopy images[J]. Photonics Research, 2021, 9(5): B168-B181.
[44] BURNS Z, LIU Z. Untrained, physics-informed neural networks for structured illumination microscopy[J]. Optics Express, 2023, 31(5): 8714-8724.
[45] WANG J, FAN J, ZHOU B, et al. Hybrid reconstruction of the physical model with the deep learning that improves structured illumination microscopy[J]. Advanced Photonics Nexus, 2023, 2(1): 016012.
[46] QIAO C, LI D, GUO Y, et al. Evaluation and development of deep neural networks for image super-resolution in optical microscopy[J]. Nature Methods, 2021, 18(2): 194-202.
[47] ZHANG H, GOODFELLOW I, METAXAS D, et al. Self-attention generative adversarial networks[C]//International Conference on Machine Learning. PMLR, 2019: 7354-7363.
[48] GULRAJANI I, AHMED F, ARJOVSKY M, et al. Improved training of wasserstein gans[J]. Advances in Neural Information Processing Systems, 2017, 30
[49] LING C, ZHANG C, WANG M, et al. Fast structured illumination microscopy via deep learning[J]. Photonics Research, 2020, 8(8): 1350-1359.
[50] CHEN J, WANG L, FENG R, et al. CycleGAN-STF: spatiotemporal fusion via cyclegan-based image generation[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(7): 5851-5865.
[51] ZHANG Q, CHEN J, LI J, et al. Deep learning-based single-shot structured illumination microscopy[J]. Optics and Lasers in Engineering, 2022, 155: 107066.
[52] ZHANG Y, FAN Q, BAO F, et al. Single-image super-resolution based on rational fractal interpolation[J]. IEEE Transactions on Image Processing, 2018, 27(8): 3782-3797.
[53] SIU W C, HUNG K W. Review of image interpolation and super-resolution[C]//Proceedings of The 2012 Asia Pacific Signal and Information Processing Association Annual Summit and Conference. IEEE, 2012: 1-10.
[54] SHI W, CABALLERO J, HUSZáR F, et al. Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network.[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 1874-1883.
[55] GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[J]. Communications of the ACM, 2020, 63(11): 139-144.
[56] KŘíŽEK P, LUKEŠ T, OVESNý M, et al. SIMToolbox: a matlab toolbox for structured illumination fluorescence microscopy[J]. Bioinformatics, 2016, 32(2): 318-320.
[57] HUANG B, BATES M, ZHUANG X. Super-resolution fluorescence microscopy[J]. Annual Review of Biochemistry, 2009, 78(1): 993-1016.
[58] 左超, 陈钱. 分辨率、超分辨率与空间带宽积拓展—从计算光学成像角度的一些思考[J]. 中国光学（中英文）, 2022, 15(6): 1105-1166.
[59] HUANG G, LIU Z, MAATEN L V D, et al. Densely connected convolutional networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 4700-4708.
[60] DING X, ZHANG X, MA N, et al. Repvgg: making vgg-style convnets great again[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 13733-13742.
[61] BARRON J T. A general and adaptive robust loss function[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019: 4331-4339.
[62] LIU C, YU S, YU M, et al. Adaptive smooth L1 loss: a better way to regress scene texts with extreme aspect ratios[C]//2021 IEEE Symposium on Computers and Communications (ISCC), IEEE, 2021: 1-7.