基于3D全卷积网络的腹部动脉CTA分割算法

doi:10.16180/j.cnki.issn1007-7820.2022.03.006

摘要/Abstract

摘要：

目前卷积神经网络已成为腹部动脉血管分割领域的研究热点,但经典的卷积网络存在分割精度低和分割血管不连续的问题。为此,文中提出了基于改进3D全卷积网络的腹部动脉血管分割算法。该方法在网络的编码路径上构造不同尺度的侧输入,并将侧输入卷积后的图像与下采样卷积后的图像进行融合,提取更多的特征信息。同时,网络中嵌入了新的多尺度特征提取模块,该模块将通道注意力与密集扩张卷积进行了融合,有效地捕获了更高层次的特征信息。对腹部动脉血管进行分割的结果表明,与其他分割方法相比,所提方法在直观性和定量性上均有提高,证明了该方法能够提升血管分割精度。

关键词: 医学图像处理, 计算机断层扫描, 腹部血管分割, 3D卷积神经网络, 密集扩张卷积, 通道注意力机制, 多尺度特征融合

Abstract:

Convolutional neural networks have become a research hotspot in the field of abdominal artery segmentation. The classic convolutional network has the problems of low segmentation accuracy and discontinuous segmentation of blood vessels. In view of these problems, this study proposes an abdominal arterial vessel segmentation algorithm based on an improved 3D full convolutional network. The side input of different scales is constructed on the encoding path of the network, and the convoluted image of side input is fused with the convoluted image of down sampling to extract more feature information. Meanwhile, a new multi-scale feature extraction module is embedded in the network. In this module, the channel attention and dense dilation convolution are introduced to capture the higher-level feature information. The experimental results on abdominal artery segmentation show that compared with other segmentation methods, the proposed method is more intuitive and quantitative, indicating that this method can improve the accuracy of blood vessel segmentation.

Key words: medical image processing, computed tomography, abdominal vascular segmentation, 3D convolution neural network, dilation convolution, channel attention mechanism, multi-scale feature fusion

中图分类号:

TP391

纪玲玉,高永彬,赵呈陆,汤先华,徐凯成,徐嘉诚. 基于3D全卷积网络的腹部动脉CTA分割算法[J]. 电子科技, 2022, 35(3): 38-44.

Lingyu JI,Yongbin GAO,Chenglu ZHAO,Xianhua TANG,Kaicheng XU,Jiacheng XU. CTA Segmentation Algorithm of Abdominal Artery Based on 3D Fully Convolutional Network[J]. Electronic Science and Technology, 2022, 35(3): 38-44.

图/表 9

图1

图2

图3

图4

图5

图6

表1

图7

图8

参考文献 32

[1]	高跃. 基于动脉期图像的腹主动脉分割研究[D]. 沈阳:东北大学, 2014.
	Gao Yue. Research of abdominal aorta segmentation based on arterial phase CT image[D]. Shenyang:Northeastern University, 2014.
[2]	孙浩. CTA 影像中血管三维分割算法的研究[D]. 武汉:中南民主大学, 2013.
	Sun Hao. The research of vessel 3D segmentation algorithm in CTA images[D].Wuhan:South-Central University for Nationalities, 2013.
[3]	郭杰龙. CT图像的腹部血管三维分割[D]. 武汉:中南民族大学, 2015.
	Guo Jielong. Abdominal aorta segmentation based on CT image[D].Wuhan:South-Central University for Nationalities, 2015.
[4]	沈烨, 方志军, 高永彬. 嵌入注意力机制的多模型融合冠脉CTA分割算法[J].计算机科学与探索, 2020(9):1602-1611.
	Shen Ye, Fang Zhijun, Gao Yongbin. Multi-model fusion of coronary CTA segmentation based on attention mechanism[J].Journal of Frontiers of Computer Science and Technology, 2020(9):1602-1611.
[5]	Soomro M H, Giunta G, Laghi A, et al. Segmenting MR images by level-set algorithms for perspective colorectal cancer diagnosis [C].Portugal:Proceedings of the European Congress on Computational Methods in Applied Sciences and Engineering, 2017.
[6]	Chen Y T. A novel approach to segmentation and measurement of medical image using level set methods[J]. Magnetic Resonance Imaging, 2017,39(4):175-193.
[7]	Wang Y, Liatsis P. An automated method for segmentation of coronary arteries in coronary CT imaging [C].London:Proceedings of the Developments in E-systems Engineering, 2010.
[8]	Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017,60(6):84-90.
[9]	Simonyan K, Zissermny A. Very deep convolutional networks for large-scale image recognition[J]. Computer Vision and Pattern Recognition, 2015,140(9):1-14.
[10]	Szegedy C, Liu W, Jia Y, et al. Going deeper with convolutions [C].Boston:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015.
[11]	Wang T, Wu D J, Coates A, et al. End-to-end text recognition with convolutional neural networks [C].Tsukuba:Proceedings of the Twenty-first International Conference on Pattern Recognition, 2012.
[12]	Shelhamer E, Long J, Darrell T. Fully convolutional networks for semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014,39(4):640-651.
[13]	Ronneberge O, Fischer P, Brox T. U-Net:Convolutional networks for biomedical image segmentation [C].Munich:Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, 2015.
[14]	Li X M, Chen H, Qi X J, et al. H-DenseUNet:Hybrid densely connected UNet for liver and tumor segmentation from CT volumes[J]. IEEE Transactions on Medical Imaging, 2018,37(12):2663-2674.
[15]	He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition [C].Piscataway:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016.
[16]	Mohammed A, Yildirim S, Farup I, et al. Y-Net:A deep convolutional neural network for polyp detection [C].Newcastle:Proceedings of the British Machine Vision Conference, 2018.
[17]	贝琛圆, 于海滨, 潘勉, 等. 基于改进U-Net网络的腺体细胞图像分割算法[J]. 电子科技, 2019,32(11):18-22.
	Bei Chenyuan, Yu Haibin, Pan Mian, et al. Gland cell image segmentation algorithm based on improved U-Net network[J]. Electronic Science and Technology, 2019,32(11):18-22.
[18]	Bai W, Suzuki H, Qin C, et al. Recurrent neural networks for aortic image sequence segmentation with sparse annotations [C].Granada:Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, 2018.
[19]	Huang W M, Huang L, Lin Z P, et al. Coronary artery segmentation by deep learning neural networks on computed tomographic coronary angiographic image[J].IEEE Engineering in Medicine and Biology Society, 2018(8):608-611.
[20]	Milletari F, Navab N, Ahmadi S A. V-Net:Fully convolutional neural networks for volumetric medical image segmentation [C].Stanford:Proceedings of the Fourth International Conference on 3D Vision, 2016.
[21]	Gibson E, Giganti F, Hu Yiping, et al. Automatic multi-organ segmentation on abdominal CT with dense V-networks[J]. IEEE Transactions on Medical Imaging, 2018,37(8):1822-1834.
[22]	Wang Y, Zhou Y Y, Shen W, et al. Abdominal multi-organ segmentation with organ-attention networks and statistical fusion[J]. Medical Image Analysis, 2019,55(2):88-102.
[23]	Holger R R, Shen C, Hirohisa O, et al. A multi-scale pyramid of 3D fully convolutional networks for abdominal multi-organ segmentation [C].Granada:Proceedings of the International on Medical Image Computing and Computer-Assisted Intervention, 2018.
[24]	Kakeya H, Okada T, Oshiro Y. 3D U-JAPA-Net:Mixture of convolutional networks for abdominal multi-organ CT segmentation [C].Granada:Proceedings of the International on Medical Image Computing and Computer-Assisted Intervention, 2018.
[25]	Casamitjana A, Cata M, Sanchez I, et al. Cascaded V-Net using ROI masks for brain tumor segmentation [C].Quebec City:Proceedings of the International on Medical Image Computing for Computer Assisted Intervention, 2017.
[26]	Zhao N N, Tong N, Ruan D, et al. Fully automated pancreas segmentation with two-stage 3D convolutional neural networks [C].Shenzhen:Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, 2019.
[27]	Chen L C, Papangreou G, Kokkinos I, et al. DeepLab:semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transaction on Pattern Analysis Machine Intelligence, 2018,40(4):834-848.
[28]	Gu Z W, Cheng J, Fu H Z, et al. CE-Net:context encoder network for 2D medical image segmentation[J]. IEEE Transactions on Medical Imaging, 2019,38(10):2281-2292.
[29]	Szegedy C, Ioffe S, Vanhoucke V, et al. Inception-v4, inception-resnet and the impact of residual connections on learning [C].San Francisco:Proceedings of the Thirty-first AAAI Conference on Artificial Intelligence, 2017.
[30]	Hu J, Shen L, Albanie S, et al. Squeeze-and-Excitation networks[J]. IEEE Transactins on Pattern Analysis Machine Intelligence, 2018,42(8):2011-2023.
[31]	刘敏, 方志军, 高永彬. 多阶段冠状动脉 CTA 三维分割算法[J]. 信号处理, 2019,35(11):1911-1918.
	Liu Min, Fang Zhijun, Gao Yongbin. Multi-stage three-dimensional segmentation algorithm for coronary CTA[J]. Journal of Signal Processing, 2019,35(11):1911-1918.
[32]	Oktay O, Schlemper J, Folgoc L, et al. Attention U-Net:learning where to look for the pancreas [C].Amsterdam:Proceedings of the Medical Imaging with Deep Learning, 2018.

方法	准确率	特异性	灵敏度	相似度
3DUNet	0.999 0	0.999 8	0.741 7	0.809 8
VNet	0.999 0	0.999 7	0.763 2	0.822 0
SE-VNet	0.999 1	0.998 9	0.753 6	0.825 9
AG-VNet	0.999 5	0.999 8	0.761 0	0.834 7
本文	0.999 7	0.999 3	0.770 2	0.847 7