多特征融合神经网络的眼底血管分割算法

doi:10.3778/j.issn.1673-9418.2104111

摘要/Abstract

摘要：

眼底毛细血管的自动监测对眼科疾病、糖尿病、心脏病等疾病的早期筛查具有重要意义。为了解决对毛细血管特征表达不精细带来的血管分割缺失问题，提出多模块融合的残差神经网络模型（MbResU-Net）。该模型利用了编码-解码网络结构。为了减少网络编码器与解码器之间的语义差距而带来的信息丢失，用非线性网络结构代替快捷连接嵌入到网络中。为了获得更多血管的细节特征，MbResU-Net提出将三块U型网络以残差方式连接，在避免丢失的前提下，最大地提取视网膜结构特征。为了保证分割质量，对图像执行预处理操作，并设计融合了代价矩阵的交叉熵损失函数来训练网络参数。对MbResU-Net与现有的眼底血管分割算法在DRIVE和CHASE DB1彩色眼底图像数据集上进行对比实验。实验表明MbResU-Net在[Sen]、[ACC]和[AUC]上优于现有方法。[Sen]为0.798 7和0.797 2，[ACC]为0.964 8和0.972 6，[AUC]为0.979 1和0.982 4。实验证明该模型在复杂曲率和小血管分割中具有有效性和鲁棒性。

关键词: 图像处理, U-Net, 残差神经网络, 多模块, 视网膜血管, 图像分割

Abstract:

In the early screening of ophthalmic diseases, diabetes, heart disease and other diseases, automatic mon-itoring of fundus capillaries occupies an important position. To solve the deficiency of blood vessel segmentation caused by the imprecise expression of capillary features, this paper proposes a multi-block residual neural network model (MbResU-Net) which uses the encoding-decoding network structure. In order to reduce the information loss caused by the semantic gap between network encoder and decoder, the model uses a nonlinear network structure instead of shortcut connections to be embedded in the network. In order to obtain the detailed features of fundus blood vessels, MbResU-Net connects three U-shaped networks in a residual block, and then extracts as much as possible low-level features of the fundus tube on the premise of avoiding information loss. To ensure the seg-mentation quality, with the proper pre-processing images to the network, it designs a cross-entropy loss function incorporating the cost matrix to train the network parameters. This paper compares MbResU-Net and the existing fundus blood vessel segmentation algorithms on the DRIVE and CHASE DB1 color fundus image datasets. Exper-iments show that MbResU-Net is superior to existing methods in Sen, ACC and AUC. The Sen is 0.7987 and 0.7972, ACC is 0.9648 and 0.9726 and AUC is 0.9791 and 0.9824. Experiments prove the effectiveness and robustness of the model in the segmentation of complex curvature and small blood vessels.

Key words: image processing, U-Net, residual neural network, multi-module, retinal vessels, image segmentation

宋姝洁, 崔振超, 陈丽萍, 陈向阳. 多特征融合神经网络的眼底血管分割算法[J]. 计算机科学与探索, 2021, 15(12): 2401-2412.

SONG Shujie, CUI Zhenchao, CHEN Liping, CHEN Xiangyang. Fundus Vessel Segmentation Algorithm Based on Multi-feature Fusion Neural Network[J]. Journal of Frontiers of Computer Science and Technology, 2021, 15(12): 2401-2412.

参考文献

[1] YU L F, DENG F H, QIN S W, et al. Classification method based on hierarchical features of fundus image[J]. Journal of Computer Applications, 2019, 39(9): 2575-2579.
余林芳, 邓伏虎, 秦少威, 等. 基于眼底图像层次特征的分类方法[J]. 计算机应用, 2019, 39(9): 2575-2579.
[2] LI D X, ZHANG Z. Improved U-Net segmentation algori-thm for the retinal blood vessel images[J]. Acta Optica Sinica, 2020, 40(10): 64-72.
[3] CHAUDHURI S, CHATTERJEE S, KATZ N, et al. Detec-tion of blood vessels in retinal images using two-dimensional matched filters[J]. IEEE Transactions on Medical Imaging, 1989, 8(3): 263-269.
[4] BEKKERS E J, DUITS R, BERENDSCHOT T, et al. A multi-orientation analysis approach to retinal vessel tracking[J]. Journal of Mathematical Imaging and Vision, 2014, 49(3): 583-610.
[5] YANG Y, HUANG S Y, RAO N. An automatic hybrid met-hod for retinal blood vessel extraction[J]. International Jour-nal of Applied Mathematics and Computer Science, 2008, 18(3): 399-407.
[6] DE J, CHENG L, ZHANG X W, et al. A graph-theoretical approach for tracing filamentary structures in neuronal and retinal images[J]. IEEE Transactions on Medical Imaging, 2016, 35(1): 257-272.
[7] ZHAO Y T, RADA L, CHEN K, et al. Automated vessel segmentation using infinite perimeter active contour model with hybrid region information with application to retinal images[J]. IEEE Transactions on Medical Imaging, 2015, 34(9): 1797-1807.
[8] RICCI E, PERFETTI R. Retinal blood vessel segmentation using line operators and support vector classification[J]. IEEE Transactions on Medical Imaging, 2007, 26(10): 1357-1365.
[9] MENG L, LIU J, CAO H, et al. Retinal vessel segmentation based on Frangi filter and Otsu algorithm[J]. Progress in Laser and Optoelectronics, 2019,?65(18): 119-125.
孟琳,?刘静,?曹慧, 等.?基于Frangi滤波器和Otsu视网膜血管分割[J]. 激光与光电子学进展,?2019,?65(18): 119-125.
[10] LISKOWSKI P, KRAWIEC K. Segmenting retinal blood vessels with deep neural networks[J]. IEEE Transactions on Medical Imaging, 2016, 35(11): 2369-2380.
[11] RONNEBERGER O, FISCHER P, BROX T. U-Net: con-volutional networks for biomedical image segmentation[C]//LNCS 9351: Proceedings of the 2015 International Conference on Medical Image Computing and Computer-Assisted Intervention. Cham: Springer, 2015: 234-241.
[12] XUE W X, LIU J X, LIU R, et al. An improved method for retinal vasular segmentation in U-Net[J]. Acta Optica Sinica, 2020, 40(12): 100-110.
薛文渲, 刘建霞, 刘然, 等. 改进U型网络的眼底视网膜血管分割方法[J]. 光学学报, 2020, 40(12): 100-110.
[13] LI D, RAHARDJA S. BSEResU-Net: an attention-based before-activation residual U-Net for retinal vessel segmen-tation[J]. Computer Methods and Programs in Biomedicine, 2021, 205: 106070.
[14] DONGYE C, MA Y Y. An improved U-Net method with high-resolution feature maps for retinal blood vessel seg-mentation[J]. Journal of Physics: Conference Series, 2021(1): 012099.
[15] HU Y T, PEI Y, LIN C, et al. Atrous residual U-Net for retinal vessel segmentation[J]. Computer Engineering and Applications, 2021,?57(7):?185-191.
胡扬涛, 裴洋, 林川, 等.?空洞残差U型网络用于视网膜血管分割[J].?计算机工程与应用,?2021,?57(7):?185-191.
[16] SZEGEDY C, IOFFE S, VANHOUCKE V, et al. Inception-v4, Inception-ResNet and the impact of residual connec-tions on learning[J]. arXiv:1602.07261, 2016.
[17] DROZDZAL M, VORONTSOV E, CHARTRAND G, et al. The importance of skip connections in biomedical image segmentation[C]//LNCS 10008: Proceedings of the 1st In-ternational Workshop on Deep Learning and Data Labeling for Medical Applications, Athens, Oct 21, 2016. Cham:Springer, 2016: 179-187.
[18] NABIL I, SOHEL R M. MultiResUNet: rethinking the U-Net architecture for multimodal biomedical images segmen-tation[J]. arXiv:1902.04049, 2019.
[19] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//Proceedings of the 2016 IEEE Con-ference on Computer Vision and Pattern Recognition, Las Vegas, Jun 27-30, 2016. Washington: IEEE Computer Society, 2016: 770-778.
[20] WANG E, JIANG Y, LI Y, et al. MFCSNet: multi-scale deep features fusion and cost-sensitive loss function-based segmentation network for remote sensing images[J]. Applied Sciences, 2019, 9(19): 4043.
[21] LI Y Y, CAI Y H, GAO X R. Retinal vessel segmentation algorithm based on hybrid phase feature[J]. Journal of Com-puter Applications, 2018, 38(7): 2083-2088.
李媛媛, 蔡轶珩, 高旭蓉. 基于融合相位特征的视网膜血管分割算法[J]. 计算机应用, 2018, 38(7): 2083-2088.
[22] ZHAO Y, ZHANG J H. Multi-scale feature fusion method for spinal X-Ray image segmentation[J]. Computer Engi-neering and Applications,?2021,?57(8):?214-219.
赵阳, 张俊华.?多尺度特征融合的脊柱X线图像分割方法[J].?计算机工程与应用,?2021,?57(8):?214-219.
[23] ZHUANG J T. LadderNet: multi-path networks based on U-Net for medical image segmentation[J]. arXiv:1810.07810, 2018.
[24] JIN Q G, MENG Z P, PHAM T D, et al. DUNet: a deforma-ble network for retinal vessel segmentation[J]. Knowledge-Based Systems, 2019, 178: 149-162.
[25] LI Q L, FENG B W, XIE L P, et al. A cross-modality lear-ning approach for vessel segmentation in retinal images[J]. IEEE Transactions on Medical Imaging, 2016, 35(1): 109-118.
[26] ORLANDO J I, PROKOFYEVA E, BLASCHKO M B. A discriminatively trained fully connected conditional ran-dom field model for blood vessel segmentation in fundus images[J]. IEEE Transactions on Biomedical Engineering, 2017, 64(1): 16-27.
[27] ALOM M Z, HASAN M, YAKOPCIC C, et al. Recurrent residual convolutional neural network based on U-Net (R2U-Net) for medical image segmentation[J]. arXiv:1802.06955, 2018.
[28] WU Y C, YONG X, YANG S, et al. Multiscale network followed network model for retinal vessel segmentation[C]//LNCS 11071: Proceedings of the 21st International Confe-rence on Medical Image Computing and Computer Assisted Intervention, Granada, Sep 16-20, 2018. Cham: Springer, 2018: 119-126.
[29] WANG B, QIU S, HE H G. Dual encoding U-Net for retinal vessel segmentation[C]//LNCS 11764: Proceedings of the 22nd International Conference on Medical Image Computing and Computer Assisted Intervention, Shenzhen, Oct 13-17, 2019. Cham: Springer, 2019: 84-92.
[30] DAN Y, LIU G R, REN M C, et al. A multi-scale feature fusion method based on U-Net for retinal vessel segmenta-tion[J]. Entropy, 2020, 22(8): 811.
[31] ATLI B, GEDIK O S. Sine-Net: a fully convolutional deep learning architecture for retinal blood vessel segmentation[J]. Engineering Science and Technology, an International Journal, 2021, 24(2): 271-283.