Abstract
Vessel segmentation is an important step for cerebrovascular disease analysis, while automatic and complete segmentation of head and neck vessels in CT angiography is a challenging problem. The reason is that vessels have diverse shapes and sizes in long and tortuous tubular-like vasculatures, as well as confounding appearance with surrounding tissues. Current deep learning-based methods often use voxel-wise segmentation, without considering the shape or connectivity of segmented vessels. In this paper, we describe vascular structures using affinity maps and construct connectivity priors of vessels so as to establish a topological connectivity loss for vessel segmentation. Specifically, a multi-head 3D U-Net is applied to predict the segmentation mask and the affinity map, and subsequently the connectivity-aware affinity map is used to refine the segmentation. In experiments, we applied our method on 72 head and neck CT angiography images. Voxel-wise and topology-relevant metrics show that the proposed method achieved superior performance than the widely used 3D nnU-Net and provided better 3D vessel visualization results.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Xu, G., Ma, M., Liu, X., Hankey, G.J.: Is there a stroke belt in china and why? Stroke 44(7), 1775–1783 (2013)
Saxena, A., Ng, E.Y.K., Lim, S.T.: Imaging modalities to diagnose carotid artery stenosis: progress and prospect. Biomed. Eng. Online. 18(1), 1–23 (2019)
Hedblom, A.: Blood vessel segmentation for neck and head computed tomography angiography (2013)
Cuisenaire, O., Virmani, S., Olszewski, M.E., Ardon, R.: Fully automated segmentation of carotid and vertebral arteries from contrast-enhanced CTA. In: Medical Imaging 2008: Image Processing, vol. 6914, p. 69143R. International Society for Optics and Photonics (2008)
Fan, F., et al.: Rapid vessel segmentation and reconstruction of head and neck angiograms using 3d convolutional neural network. Nat. Commun. 11(1), 1–12 (2020)
Xu, R., Liu, T., Ye, X., Lin, L., Chen, Y.-W.: Boosting connectivity in retinal vessel segmentation via a recursive semantics-guided network. In: Martel, A.L., et al. (eds.) MICCAI 2020. LNCS, vol. 12265, pp. 786–795. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-59722-1_76
Paetzold, J.C., et al.: clDice-a novel connectivity-preserving loss function for vessel segmentation. In: Medical Imaging Meets NeurIPS 2019 Workshop (2019)
Funke, J., et al.: Large scale image segmentation with structured loss based deep learning for connectome reconstruction. IEEE Trans. Pattern Anal. Mach. Intell. 41(7), 1669–1680 (2018)
Hu, X., Fuxin, L., Samaras, D., Chen, C.: Topology-preserving deep image segmentation. arXiv preprint arXiv:1906.05404 (2019)
Qin, Y., et al.: AirwayNet: a voxel-connectivity aware approach for accurate airway segmentation using convolutional neural networks. In: Shen, D., et al. (eds.) MICCAI 2019. LNCS, vol. 11769, pp. 212–220. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-32226-7_24
Kruskal, J.B.: On the shortest spanning subtree of a graph and the traveling salesman problem. Proc. Am. Math. Soc. 7(1), 48–50 (1956)
Çiçek, Ö., Abdulkadir, A., Lienkamp, S.S., Brox, T., Ronneberger, O.: 3D U-net: learning dense volumetric segmentation from sparse annotation. In: Ourselin, S., Joskowicz, L., Sabuncu, M.R., Unal, G., Wells, W. (eds.) MICCAI 2016. LNCS, vol. 9901, pp. 424–432. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46723-8_49
Isensee, F., Jaeger, P.F., Kohl, S.A.A., Petersen, J., Maier-Hein, K.H.: nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat. Methods 18(2), 203–211 (2021)
Acknowledgements
This work was supported in part by the National Key Research and Development Program of China under Grant 2018YFC0116400.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 Springer Nature Switzerland AG
About this paper
Cite this paper
Yao, L. et al. (2022). Head and Neck Vessel Segmentation with Connective Topology Using Affinity Graph. In: Lian, C., Cao, X., Rekik, I., Xu, X., Cui, Z. (eds) Machine Learning in Medical Imaging. MLMI 2022. Lecture Notes in Computer Science, vol 13583. Springer, Cham. https://doi.org/10.1007/978-3-031-21014-3_24
Download citation
DOI: https://doi.org/10.1007/978-3-031-21014-3_24
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-21013-6
Online ISBN: 978-3-031-21014-3
eBook Packages: Computer ScienceComputer Science (R0)