基于深度学习与傅里叶变换机制的医学图像分割算法的研究

医学图像分割技术在定位和测量人体组织或病变的位置和大小方面发挥着重 要作用，可用于凸显图像中的解剖结构或病理变化。深度学习算法能够准确地识 别和分割医学图像中的特定结构，如肿瘤、器官和其他病理特征。这种高精度的 图像分析有助于提高疾病的检测率和诊断的准确性。本文对提升医学图像分割性 能的深度学习网络进行了深入研究，提出了一种基于 U 形结构的轻量级网络，结 合傅里叶变换、注意力机制和残差结构的优势，提高医学图像分割的准确性同时 降低网络训练对计算资源的要求。主要研究内容如下： （1）提出轻量级分割网络 FRUNet 提出了一种轻量化网络架构，这一架构运用多项网络结构优势，提升了图像分 割的精确度。其创新之处主要体现在整合了傅里叶通道注意力模块，该模块通过 在频率域对图像进行分析处理，有效地弥合了高频与低频成分之间的差异，增强 了对细节和边缘部分的识别能力。此外，该模型基于 U-Net 架构进行改进，U-Net 的跳跃连接能够同时维持图像的广泛上下文信息和关键细节特征，从而在解析复 杂的医学图像时保持原有图像信息的完整性。而与残差网络的结合不仅加快了模 型学习过程的收敛速度，还提升了训练的效率，使模型在深层次网络训练时也能 有效避免信息丢失，解决了梯度消失或爆炸等常见问题，从而提升了整体的稳定 性和鲁棒性。各项技术的结合，网络得以在较少计算量下满足网络对特征的提取， 在提升分割性能的同时大大降低网络参数。 （2）FRUNet 网络结构与分割性能的多角度评估 经过对网络结构和参数的调整及验证，本研究成功开发了 FRUNet 这一网络 架构。通过消融实验详细分析了网络中每个组件的贡献，并展示了它们如何协同 工作以实现最佳性能。使用热力图分析直观地展现了网络各部分对图像处理时注 意力聚焦区域与关注重点的影响。为全面评估该网络在医学图像分割上的表现，我 们对细胞类型、放大倍率和成像方式等方面具有显著差异的 3 种医学图像进行了 细致的分割测试。同时引入标准差分析和 T 检验从统计学角度验证了网络的性能 和各组成部分的重要性。

[RÖNTGEN W C. On a new kind of rays[J]. Science, 1896, 3(59): 227-231.
[2] GOLDMAN L W. Principles of CT and CT technology[J]. Journal of Nuclear Medicine Technology, 2007, 35(3): 115-128.
[3] PLEWES D B, KUCHARCZYK W. Physics of MRI: a primer[J]. Journal of Magnetic Resonance Imaging, 2012, 35(5): 1038-1054.
[4] HESS S, BLOMBERG B A, ZHU H J, et al. The pivotal role of FDG-PET/CT in modernmedicine[J]. Academic Radiology, 2014, 21(2): 232-249.
[5] MCAULIFFE M J, LALONDE F M, MCGARRY D, et al. Medical image processing, analysis and visualization in clinical research[C]//Proceedings 14th IEEE Symposium on ComputerBased Medical Systems. CBMS 2001. IEEE, 2001: 381-386.
[6] KIDWELL C S, CHALELA J A, SAVER J L, et al. Comparison of MRI and CT for detectionof acute intracerebral hemorrhage[J]. Jama, 2004, 292(15): 1823-1830.
[7] SCHUSTER D M. Clinical utility of PET scanning in breast cancer management[J]. Am. J.Hematol. Oncol, 2015, 11: 20-25.
[8] ZHOU S K, GREENSPAN H, DAVATZIKOS C, et al. A review of deep learning in medicalimaging: Imaging traits, technology trends, case studies with progress highlights, and futurepromises[J]. Proceedings of the IEEE, 2021, 109(5): 820-838.
[9] LITJENS G, KOOI T, BEJNORDI B E, et al. A survey on deep learning in medical imageanalysis[J]. Medical Image Analysis, 2017, 42: 60-88.
[10] RANJBARZADEH R, CAPUTO A, TIRKOLAEE E B, et al. Brain tumor segmentation of MRIimages: A comprehensive review on the application of artificial intelligence tools[J]. Computersin Biology and Medicine, 2023, 152: 106405.
[11] RONNEBERGER O, FISCHER P, BROX T. U-net: Convolutional networks for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III 18. Springer, 2015: 234-241.
[12] JIANG H, DIAO Z, SHI T, et al. A review of deep learning-based multiple-lesion recognitionfrom medical images: classification, detection and segmentation[J]. Computers in Biology andMedicine, 2023: 106726.
[13] LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553): 436-444.
[14] BEAUDOIN N, BEAUCHEMIN S S. An accurate discrete Fourier transform for image processing[C]//2002 International Conference on Pattern Recognition: volume 3. IEEE, 2002:935-939.53参考文献
[15] POPA C A, CERNĂZANU-GLĂVAN C. Fourier transform-based image classification usingcomplex-valued convolutional neural networks[C]//Advances in Neural Networks–ISNN 2018:15th International Symposium on Neural Networks, ISNN 2018, Minsk, Belarus, June 25–28,2018, Proceedings 15. Springer, 2018: 300-309.
[16] HAN Y, HONG B W. Deep learning based on fourier convolutional neural network incorporating random kernels[J]. Electronics, 2021, 10(16): 2004.
[17] QIAO C, LI D, LIU Y, et al. Rationalized deep learning super-resolution microscopy for sustained live imaging of rapid subcellular processes[J]. Nature Biotechnology, 2023, 41(3): 367-377.
[18] CASTLEMAN K R. Digital image processing[M]. Prentice Hall Press, 1996.
[19] PHAM D L, XU C, PRINCE J L. Current methods in medical image segmentation[J]. AnnualReview of Biomedical Engineering, 2000, 2(1): 315-337.
[20] CANNY J. A computational approach to edge detection[J]. IEEE Transactions on PatternAnalysis and Machine Intelligence, 1986(6): 679-698.
[21] ZITOVA B, FLUSSER J. Image registration methods: a survey[J]. Image and Vision Computing, 2003, 21(11): 977-1000.
[22] JAIN A K, DUBES R C. Algorithms for clustering data[M]. Prentice-Hall, Inc., 1988.
[23] CHEN X, WANG X, ZHANG K, et al. Recent advances and clinical applications of deeplearning in medical image analysis[J]. Medical Image Analysis, 2022, 79: 102444.
[24] LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2015: 3431-3440.
[25] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//Proceedingsof the IEEE Conference on Computer Vision and Pattern Recognition. 2016: 770-778.
[26] HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 7132-7141.
[27] WOO S, PARK J, LEE J Y, et al. Cbam: Convolutional block attention module[C]//Proceedingsof the European conference on computer vision (ECCV). 2018: 3-19.
[28] ATIYA S U, RAMESH N. Deep Lab v3+: A novel deep learning model for accurate andefficient GTV segmentation and classification in NSCLC imaging[J]. International Journal ofIntelligent Systems and Applications in Engineering, 2024, 12(1s): 393-410.
[29] XIAO H, LI L, LIU Q, et al. Transformers in medical image segmentation: A review[J].Biomedical Signal Processing and Control, 2023, 84: 104791.
[30] GÜVEN S A, TALU M F. Brain MRI high resolution image creation and segmentation withthe new GAN method[J]. Biomedical Signal Processing and Control, 2023, 80: 104246.
[31] KAZEROONI A F, KHALILI N, LIU X, et al. The brain tumor segmentation (BRATS) challenge 2023: Focus on pediatrics (CBTN-CONNECT-DIPGR-ASNR-MICCAI BraTS-PEDs)[A]. 2023.
[32] BILIC P, CHRIST P, LI H B, et al. The liver tumor segmentation benchmark (lits)[J]. MedicalImage Analysis, 2023, 84: 102680.54参考文献
[33] SHIN H, KIM H, KIM S, et al. SDC-UDA: volumetric unsupervised domain adaptationframework for slice-direction continuous cross-modality medical image segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023:7412-7421.
[34] SHARMA N, GUPTA S, KOUNDAL D, et al. U-Net model with transfer learning model as abackbone for segmentation of gastrointestinal tract[J]. Bioengineering, 2023, 10(1): 119.
[35] GAO S, ZHOU H, GAO Y, et al. BayeSeg: Bayesian modeling for medical image segmentationwith interpretable generalizability[J]. Medical Image Analysis, 2023, 89: 102889.
[36] GOMES T, MATIAS D, CAMPOS A, et al. A survey on ground segmentation methods forautomotive LiDAR sensors[J]. Sensors, 2023, 23(2): 601.
[37] AGARWAL M, GUPTA S K, BISWAS K. Development of a compressed FCN architecture forsemantic segmentation using particle swarm optimization[J]. Neural Computing and Applications, 2023, 35(16): 11833-11846.
[38] ZHANG J, CUI Z, SHI Z, et al. A robust and efficient AI assistant for breast tumor segmentationfrom DCE-MRI via a spatial-temporal framework[J]. Patterns, 2023, 4(9).
[39] BRIGHAM E O. The fast Fourier transform and its applications[M]. Prentice-Hall, Inc., 1988.
[40] KNOLL F, HAMMERNIK K, ZHANG C, et al. Deep-learning methods for parallel magneticresonance imaging reconstruction: A survey of the current approaches, trends, and issues[J].IEEE Signal Processing Magazine, 2020, 37(1): 128-140.
[41] QIAO C, LI D, GUO Y, et al. Evaluation and development of deep neural networks for imagesuper-resolution in optical microscopy[J]. Nature Methods, 2021, 18(2): 194-202.
[42] XU K, QIN M, SUN F, et al. Learning in the frequency domain[C]//Proceedings of theIEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020: 1740-1749.
[43] GHOSH A, CHELLAPPA R. Deep feature extraction in the DCT domain[C]//2016 23rd International Conference on Pattern Recognition (ICPR). IEEE, 2016: 3536-3541.
[44] QIN Z, ZHANG P, WU F, et al. Fcanet: Frequency channel attention networks[C]//Proceedingsof the IEEE/CVF International Conference on Computer Vision. 2021: 783-792.
[45] HUBEL D H, WIESEL T N. Receptive fields, binocular interaction and functional architecturein the cat’s visual cortex[J]. The Journal of Physiology, 1962, 160(1): 106.
[46] FUKUSHIMA K. Neocognitron: A self-organizing neural network model for a mechanismof pattern recognition unaffected by shift in position[J]. Biological Cybernetics, 1980, 36(4):193-202.
[47] LECUN Y, BOTTOU L, BENGIO Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.
[48] RUMELHART D E, HINTON G E, WILLIAMS R J. Learning representations by backpropagating errors[J]. Nature, 1986, 323(6088): 533-536.
[49] HINTON G E, OSINDERO S, TEH Y W. A fast learning algorithm for deep belief nets[J].Neural Computation, 2006, 18(7): 1527-1554.
[50] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Imagenet classification with deep convolutional neural networks[J]. Advances in Neural Information Processing Systems, 2012, 25.55参考文献
[51] BASHA S S, DUBEY S R, PULABAIGARI V, et al. Impact of fully connected layers onperformance of convolutional neural networks for image classification[J]. Neurocomputing,2020, 378: 112-119.
[52] HE J, LI L, XU J, et al. ReLU deep neural networks and linear finite elements[A]. 2018.
[53] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale imagerecognition[A]. 2014: 1409.1556.
[54] SZEGEDY C, LIU W, JIA Y, et al. Going deeper with convolutions[C]//Proceedings of theIEEE Conference on Computer Vision and Pattern Recognition. 2015: 1-9.
[55] 李玉慧, 梁创学, 李军. 基于 Group-Depth U-Net 的电子显微图像中神经元结构分割[J]. 中国医学物理学杂志, 2020, 37(6): 6.
[56] TRAN S T, CHENG C H, LIU D G. A multiple layer U-Net, U n-Net, for liver and liver tumorsegmentation in CT[J]. IEEE Access, 2020, 9: 3752-3764.
[57] HWANG H, REHMAN H Z U, LEE S. 3D U-Net for skull stripping in brain MRI[J]. AppliedSciences, 2019, 9(3): 569.
[58] PARK D, KIM K, YOUNG CHUN S. Efficient module based single image super resolution formultiple problems[C]//Proceedings of the IEEE Conference on Computer Vision and PatternRecognition Workshops. 2018: 882-890.
[59] XU C, LU C, LIANG X, et al. Multi-loss regularized deep neural network[J]. IEEE Transactionson Circuits and Systems for Video Technology, 2015, 26(12): 2273-2283.
[60] RUBY U, YENDAPALLI V. Binary cross entropy with deep learning technique for imageclassification[J]. Int. J. Adv. Trends Comput. Sci. Eng, 2020, 9(10): 2278-3091.
[61] ZHAO R, QIAN B, ZHANG X, et al. Rethinking dice loss for medical image segmentation[C]//2020 IEEE International Conference on Data Mining (ICDM). IEEE, 2020: 851-860.
[62] KURMI Y, CHAURASIA V, KAPOOR N. Design of a histopathology image segmentationalgorithm for CAD of cancer[J]. Optik, 2020, 218: 164636.
[63] FAN X, ZHU Q, TU P, et al. A review of advances in image-guided orthopedic surgery[J].Physics in Medicine & Biology, 2023, 68(2): 02TR01.
[64] SZTILKOVICS M, GERECSEI T, PETER B, et al. Single-cell adhesion force kinetics of cellpopulations from combined label-free optical biosensor and robotic fluidic force microscopy[J]. Scientific Reports, 2020, 10(1): 61.
[65] LIN S, SCHORPP K, ROTHENAIGNER I, et al. Image-based high-content screening in drugdiscovery[J]. Drug Discovery Today, 2020, 25(8): 1348-1361.
[66] GREENWALD N F, MILLER G, MOEN E, et al. Whole-cell segmentation of tissue imageswith human-level performance using large-scale data annotation and deep learning[J]. NatureBiotechnology, 2022, 40(4): 555-565.
[67] GÓMEZ-DE MARISCAL E, GARCÍA-LÓPEZ-DE HARO C, OUYANG W, et al. DeepImageJ: A user-friendly environment to run deep learning models in ImageJ[J]. Nature Methods,2021, 18(10): 1192-1195.
[68] CAICEDO J C, GOODMAN A, KARHOHS K W, et al. Nucleus segmentation across imagingexperiments: the 2018 Data Science Bowl[J]. Nature Methods, 2019, 16(12): 1247-1253.56参考文献
[69] SIRINUKUNWATTANA K, PLUIM J P, CHEN H, et al. Gland segmentation in colon histologyimages: The glas challenge contest[J]. Medical Image Analysis, 2017, 35: 489-502.
[70] KUMAR N, VERMA R, ANAND D, et al. A multi-organ nucleus segmentation challenge[J].IEEE Transactions on Medical Imaging, 2019, 39(5): 1380-1391.
[71] ABADI M, BARHAM P, CHEN J, et al. {TensorFlow}: a system for {Large-Scale} machine learning[C]//12th USENIX Symposium on Operating Systems Design and Implementation (OSDI 16). 2016: 265-283.
[72] KETKAR N, KETKAR N. Introduction to keras[J]. Deep Learning with Python: a Hands-onIntroduction, 2017: 97-111.
[73] KINGMA D P, BA J. Adam: A method for stochastic optimization[A]. 2014.
[74] ZHOU Z, SIDDIQUEE M M R, TAJBAKHSH N, et al. Unet++: Redesigning skip connectionsto exploit multiscale features in image segmentation[J]. IEEE Transactions on Medical Imaging,2019, 39(6): 1856-1867.
[75] JHA D, SMEDSRUD P H, RIEGLER M A, et al. Resunet++: An advanced architecture formedical image segmentation[C]//2019 IEEE International Symposium on Multimedia (ISM).IEEE, 2019: 225-2255.
[76] CHEN L C, ZHU Y, PAPANDREOU G, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[C]//Proceedings of the European Conference on Computer Vision (ECCV). 2018: 801-818.
[77] JHA D, RIEGLER M A, JOHANSEN D, et al. Doubleu-net: A deep convolutional neuralnetwork for medical image segmentation[C]//2020 IEEE 33rd International Symposium onComputer-based Medical Systems (CBMS). IEEE, 2020: 558-564.
[78] JHA D, ALI S, TOMAR N K, et al. Real-time polyp detection, localization and segmentationin colonoscopy using deep learning[J]. IEEE Access, 2021, 9: 40496-40510.
[79] SRIVASTAVA A, JHA D, CHANDA S, et al. MSRF-Net: a multi-scale residual fusion networkfor biomedical image segmentation[J]. IEEE Journal of Biomedical and Health Informatics,2021, 26(5): 2252-2263.
[80] VALANARASU J M J, OZA P, HACIHALILOGLU I, et al. Medical transformer: Gated axialattention for medical image segmentation[C]//Medical Image Computing and Computer Assisted Intervention–MICCAI 2021: 24th International Conference, Strasbourg, France, September 27–October 1, 2021, Proceedings, Part I 24. Springer, 2021: 36-46.
[81] XIAO X, LIAN S, LUO Z, et al. Weighted res-unet for high-quality retina vessel segmentation[C]//2018 9th International Conference on Information Technology in Medicine and Education(ITME). IEEE, 2018: 327-331