全局特征及多层次特征聚合的冠脉分割算法

doi:10.3778/j.issn.1673-9418.2005007

摘要/Abstract

摘要：

冠脉计算机断层扫描血管造影（CTA）图像分割在辅助医生判断血管堵塞、血管疾病诊断等许多实际应用中发挥重要作用。针对CTA图像中存在大量噪声和FCN、U-Net、V-Net等经典深度学习算法分割结果不细腻的问题，提出了全局特征及多层次特征聚合网络。这种新型的网络由全局特征模块、特征融合与V形细化多层次特征聚合模块以及深度监督三部分组成。全局特征模块综合早期和后期特征信息，在融合丰富的细节和语义信息基础上实现对原始CTA图像过滤操作，生成基础特征。细化V形模块在基础特征的基础上生成不同层次的细化特征图，通过聚合不同层次的细化特征图，得到精准冠脉分割图像。此外，在每一个细化V形模块之后加入深度监督机制来避免梯度消失的问题。对提出的方法进行了定量与定性的分析，结果表明，该方法优于主流基线。消融实验也证明了每个模块的有效性。

关键词: 冠脉计算机断层扫描血管造影（CTA）分割, 全局特征, 细化V形, 多层次特征聚合, 深度监督

Abstract:

Coronary computed tomographic angiography (CTA) image segmentation plays an important role in many practical applications, such as assisting doctors to judge vascular occlusion, vascular disease diagnosis, etc. In view of the large amount of noise in CTA images and not delicate segmentation results of traditional deep learning algorithms (including FCN, U-Net, V-Net, etc.), this paper proposes a global feature and multi-level feature aggregation network, which includes three network models, global feature module, feature fusion and refined V-shape multi-level feature aggregation module, and deep supervision module. The global feature module can filter the original CTA image and generate the basic features by integrating the early and later feature information and integrating the rich details and semantic information. The refined V-shape module generates refined feature maps of different levels on the basis of basic features, and obtains accurate coronary segmentation images by aggregating the refined feature maps of different levels. In addition, a deep supervision mechanism is added after each refined V-shape module to avoid the problem of gradient disappearing. The results show that the proposed method is superior to the mainstream baseline intuitively and quantitively. The ablation experiments also prove the effectiveness of each module.

Key words: coronary computed tomographic angiography (CTA) segmentation, global feature, refined V-shape, multi-level feature aggregation, deep supervision

顾佳, 方志军, 田方正. 全局特征及多层次特征聚合的冠脉分割算法[J]. 计算机科学与探索, 2021, 15(5): 958-970.

GU Jia, FANG Zhijun, TIAN Fangzheng. Global Feature and Multi-level Feature Aggregation Segmentation Algorithm for Coronary[J]. Journal of Frontiers of Computer Science and Technology, 2021, 15(5): 958-970.

参考文献

[1] WANG H, ABAJOBIR A A, ABATE K H, et al. Global, re-gional, and national under-5 mortality, adult mortality, age-specific mortality, and life expectancy, 1970-2016: a syste-matic analysis for the global burden of disease study 2016[J]. The Lancet, 2017, 390(10100): 1084-1150.
[2] CURY R C, ABBARA S, ACHENBACH S, et al. Coronary artery disease-reporting and data system (CAD-RADS): an expert consensus document of SCCT, ACR and NASCI: endorsed by the ACC[J]. JACC: Cardiovascular Imaging, 2016, 9(9): 1099-1113.
[3] DEWEY M, RUTSCH W, SCHNAPAUFF D, et al. Coronary artery stenosis quantification using multislice computed to-mography[J]. Investigative Radiology, 2007, 42(2): 78-84.
[4] LI L H, HUANG Y S, YANG R Q, et al. Segmentation of coronary artery from dual-source CT images[J]. Journal of Clinical Rehabilitative Tissue Engineering Research, 2012, 16(39): 7298-7301.
黎丽华, 黄岳山, 杨荣骞, 等. 基于双源CT图像的冠状动脉分割[J]. 中国组织工程研究, 2012, 16(39): 7298-7301.
[5] CAI K, YANG R Q, LI L H, et al. A semi-automatic coronary artery segmentation framework using mechanical simulation[J]. Journal of Medical Systems, 2015, 39(10): 1-7.
[6] HERNANDEZ M, FRANGI A F. Non-parametric geodesic active regions: method and evaluation for cerebral aneurysms segmentation in 3DRA and CTA[J]. Medical Image Analysis, 2007, 11(3): 224-241.
[7] LACOSTE C, FINET G, MAGNIN I E. Coronary tree extrac-tion from X-ray angiograms using marked point processes[C]//Proceedings of the 2006 IEEE International Sympo-sium on Biomedical Imaging: From Nano to Macro, Arlin-gton, Apr 6-9, 2006. Piscataway: IEEE, 2006: 157-160.
[8] ZHAO J, GONG W K, JIANG S Z, et al. Automatic segmen-tation and reconstruction of coronary arteries based on sphere model and Hessian matrix using CCTA images[J]. Journal of Physics: Conference Series, 2019, 1213(4): 042049.
[9] KRIZHEVSKY A, SUTSKEVER I, HINTON G E, et al. ImageNet classification with deep convolutional neural net-works[C]//Proceedings of the 26th Neural Information Pro-cessing Systems, Lake Tahoe, Dec 3-6, 2012. Red Hook: Curran Associates, 2012: 1097-1105.
[10] WANG Z, YIN Y X, SHI J P, et al. Zoom-in-net: deep mining lesions for diabetic retinopathy detection[C]//LNCS 10435: Proceedings of the 20th International Conference on Medical Image Computing and Computer-Assisted Interven-tion, Quebec, Sep 11-13, 2017. Berlin, Heidelberg: Springer, 2017: 267-275.
[11] CIRESAN D C, GIUSTI A, GAMBARDELLA L M, et al. Deep neural networks segment neuronal membranes in elec-tron microscopy images[C]//Proceedings of the 25th Inter-national Conference on Neural Information Processing Sys-tems, Lake Tahoe, Dec 3-6, 2012. Red Hook: Curran Asso-ciates, 2012: 2843-2851.
[12] LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[C]//Proceedings of the 33rd IEEE Conference on Computer Vision and Pattern Re-cognition, Boston, Jun 7-12, 2015. Washington: IEEE Com-puter Society, 2015: 3431-3440.
[13] RONNEBERGER O, FISCHER P, BROX T, et al. U-Net: convolutional networks for biomedical image segmentation[C]//LNCS 9351: Proceedings of the 18th International Con-ference on Medical Image Computing and Computer-Assisted Intervention, Munich, Oct 5-9, 2015. Berlin, Heidelberg: Springer, 2015: 234-241.
[14] FU H Z, CHENG J, XU Y W, et al. Joint optic disc and cup segmentation based on multi-label deep network and polar transformation[J]. IEEE Transactions on Medical Imaging, 2018, 37(7): 1597-1605.
[15] GIBSON E, GIGANTI F, HU Y, et al. Automatic multi-organ segmentation on abdominal CT with dense V-Networks[J]. IEEE Transactions on Medical Imaging, 2018, 37(8): 1822-1834.
[16] GU Z, CHENG J, FU H, et al. CE-Net: context encoder net-work for 2D medical image segmentation[J]. IEEE Trans-actions on Medical Imaging, 2019, 38(10): 2281-2292.
[17] ROTH H R, ODA H, HAYASHI Y, et al. Hierarchical 3D fully convolutional networks for multi-organ segmentation[J]. arXiv:1704.06382, 2017.
[18] ?I?EK ?, ABDULKADIR A, LIENKAMP S S, et al. 3D U-Net: learning dense volumetric segmentation from sparse annotation[C]//LNCS 9901: Proceedings of the 19th Inter-national Conference on Medical Image Computing and Computer-Assisted Intervention, Athens, Oct 17-21, 2016. Berlin, Heidelberg: Springer, 2016: 424-432.
[19] MILLETARI F, NAVAB N, AHMADI S A. V-Net: fully convolutional neural networks for volumetric medical image segmentation[C]//Proceedings of the 4th International Con-ference on 3D Vision, Stanford, Oct 25-28, 2016. Wash-ington: IEEE Computer Society, 2016: 565-571.
[20] MA G, YANG J, HUANG Y, et al. A novel automatic coro-nary artery segmentation method based on region growing with annular and spherical sector partition[J]. Journal of Me-dical Imaging and Health Informatics, 2019, 9(1): 148-152.
[21] GE S, SHI Z F, PENG G M, et al. Two-steps coronary artery segmentation algorithm based on improved level set model in combination with weighted shape-prior constraints[J]. Journal of Medical Systems, 2019, 43(7): 1-10.
[22] CHEN Y C, LIN Y C, WANG C P, et al. Coronary artery segmentation in cardiac CT angiography using 3D multi-channel U-net[J]. arXiv:1907.12246, 2019.
[23] WOLTERINK J M, LEINER T, I?GUM I. Graph convolu-tional networks for coronary artery segmentation in cardiac CT angiography[C]//LNCS 11849: Proceedings of the 1st International Workshop on Graph Learning in Medical Imaging, Shenzhen, Oct 17, 2019. Berlin, Heidelberg: Spr-inger, 2019: 62-69.
[24] FU Y, GUO B J, LEI Y, et al. Mask R-CNN based coronary artery segmentation in coronary computed tomography ang-iography[C]//Proceedings of the 1st International Confer-ence on Medical Imaging and Computer-Aided Diagnosis, Oxford, Jan 20-21, 2020. Berlin, Heidelberg: Springer, 2020, 11314: 113144F.
[25] KONG B, WANG X, BAI J, et al. Learning tree-structured representation for 3D coronary artery segmentation[J]. Com-puterized Medical Imaging and Graphics, 2020, 80: 101688.
[26] BLAIECH A G, MANSOUR A, KERKENI A, et al. Impact of enhancement for coronary artery segmentation based on deep learning neural network[C]//LNCS 11868: Proceedings of the 9th Iberian Conference on Pattern Recognition and Image Analysis, Madrid, Jul 1-4, 2019. Berlin, Heidelberg: Springer, 2019: 260-272.
[27] WANG L, LIANG D, YIN X, et al. Coronary artery seg-mentation in angiographic videos using A 3D-2D CE-Net[J]. arXiv:2003.11851, 2020.
[28] DOU Q, CHEN H, JIN Y M, et al. 3D deeply supervised network for automatic liver segmentation from CT volumes[C]//Proceedings of the 19th International Conference on Medical Image Computing and Computer-Assisted Interven-tion, Istanbul, Oct 17-21, 2016. Berlin, Heidelberg: Springer, 2016: 149-157.
[29] ZHENG Y C, GENG C, SONG S, et al. Novel 3D coronary artery extraction method using hierarchical clustering of spherical operator[J]. Journal of Image and Graphics, 2014, 19(8): 1219-1227.
郑永昌, 耿辰, 宋爽, 等. 应用球形算子层次聚类的3维冠脉跟踪提取[J]. 中国图象图形学报, 2014, 19(8): 1219-1227.
[30] LIU B, GU L, LU F. Unsupervised ensemble strategy for retinal vessel segmentation[C]//LNCS 11764: Proceedings of the 22nd International Conference on Medical Image Com-puting and Computer-Assisted Intervention, Shenzhen, Oct 13-17, 2019. Berlin, Heidelberg: Springer, 2019: 111-119.
[31] YU Q H, YANG D, ROTH H, et al. C2FNAS: coarse-to-fine neural architecture search for 3D medical image seg-mentation[C]//Proceedings of the 2020 IEEE/CVF Confer-ence on Computer Vision and Pattern Recognition, Seattle, Jun 13-19, 2020. Piscataway: IEEE, 2020: 4125-4134.
[32] FANG Y Q, CHEN C, YUAN Y X, et al. Selective feature aggregation network with area-boundary constraints for polyp segmentation[C]//LNCS 11764: Proceedings of the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, Shenzhen, Oct 13-17, 2019. Berlin, Heidelberg: Springer, 2019: 302-310.
[33] YU F, ZHAO J, GONG Y J, et al. Annotation-free cardiac vessel segmentation via knowledge transfer from retinal images[C]//LNCS 11765: Proceedings of the 22nd Interna-tional Conference on Medical Image Computing and Com-puter-Assisted Intervention, Shenzhen, Oct 13-17, 2019. Berlin, Heidelberg: Springer, 2019: 714-722.
[34] ZHANG S H, FU H Z, YAN Y G, et al. Attention guided network for retinal image segmentation[C]//LNCS 11764: Proceedings of the 22nd International Conference on Med-ical Image Computing and Computer-Assisted Intervention, Shenzhen, Oct 13-17, 2019. Berlin, Heidelberg: Springer, 2019: 797-805.
[35] HUANG C, HAN H, YAO Q S, et al. 3D U2-Net: a 3D universal U-Net for multi-domain medical image segmen-tation[C]//LNCS 11765: Proceedings of the 22nd Interna-tional Conference on Medical Image Computing and Com-puter-Assisted Intervention, Shenzhen, Oct 13-17, 2019. Berlin, Heidelberg: Springer, 2019: 291-299.