Abstract
Accurate whole-heart segmentation plays an important role in the surgical planning for heart defects such as congenital heart disease (CHD). In this work, we propose a deep learning method for automatic whole-heart segmentation in cardiac magnetic resonance (CMR) images with CHD. First, we start with a 3D fully convolutional network (3D FCN) in order to ensure an efficient voxel-wise labeling. Then we introduce dilated convolutional layers (3D-HOL layers) into the baseline model to expand its receptive field, so as to make better use of the spatial information. Last, we employ deeply-supervised pathways to accelerate training and exploit multi-scale information. We evaluate the proposed method on 3D CMR images from the dataset of the HVSMR 2016 Challenge. The results of controlled experiments demonstrate the efficacy of the proposed 3D-HOL layers and deeply-supervised pathways. We achieve an average Dice score of 80.1% in training (5-fold cross-validation) and 69.5% in testing.
J. Li and R. Zhang—Joint first authors.
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
Notes
- 1.
Other combination operators such as summation or max fusion need more memory to store the new volumes, whereas by concatenation we can directly do the processing on the existed volumes.
- 2.
- 3.
During the training phase, we noticed that the time point at which the loss of the 3D-FCN-HOL network started to decrease was later than the baseline’s, and the descent speed was also lower, which indicates that the network with 3D-HOL layers is more difficult to train.
References
Chen, L.C., Papandreou, G., Kokkinos, I., Murphy, K., Yuille, A.L.: Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs. arXiv preprint arXiv:1412.7062 (2014)
Dou, Q., Chen, H., Yu, L., Zhao, L., Qin, J., Wang, D., Mok, V.C., Shi, L., Heng, P.A.: Automatic detection of cerebral microbleeds from MR images via 3D convolutional neural networks. IEEE Trans. Med. Imaging 35, 1182–1195 (2016)
He, K., Zhang, X., Ren, S., Sun, J.: Delving deep into rectifiers: surpassing human-level performance on imagenet classification. In: 28th IEEE International Conference on Computer Vision, pp. 1026–1034. IEEE Press, New York (2015)
Ioffe, S., Szegedy, C.: Batch normalization: accelerating deep network training by reducing internal covariate shift. arXiv preprint arXiv:1502.03167 (2015)
Kamnitsas, K., Ledig, C., Newcombe, V.F., Simpson, J.P., Kane, A.D., Menon, D.K., Rueckert, D., Glocker, B.: Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation. arXiv preprint arXiv:1603.05959
Lee, C.Y., Xie, S., Gallagher, P., Zhang, Z., Tu, Z.: Deeply-supervised nets. In: 18th International Conference on Artificial Intelligence and Statistics, p. 6, San Diego (2015)
Long, J., Shelhamer, E., Darrell, T.: Fully convolutional networks for semantic segmentation. In: 28th IEEE Conference on Computer Vision and Pattern Recognition, pp. 3431–3440. IEEE Press, New York (2015)
Pace, D.F., Dalca, A.V., Geva, T., Powell, A.J., Moghari, M.H., Golland, P.: Interactive whole-heart segmentation in congenital heart disease. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 80–88. Springer, Heidelberg (2015). doi:10.1007/978-3-319-24574-4_10
Zhuang, X.: Challenges and methodologies of fully automatic whole heart segmentation: a review. J. Healthc. Eng. 4, 371–408 (2013)
Zhuang, X., Shen, J.: Multi-scale patch and multi-modality atlases for whole heart segmentation of MRI. Med. Image Anal. 31, 77–87 (2016)
Acknowledgements
The work described in this paper was partially supported by a grant from the Science, Technology and Innovation Commission of Shenzhen Municipality (No.: CXZZ20140606164105361), a grant from Technology and Business Development Fund (No.: TBF15MED004), a grant from the Innovation and Technology Commission (No.: ITS/293/14FP), and grants from the Research Grants Council of the Hong Kong Special Administrative Region, China (No.: CUHK 416712, and CUHK 14113214), and a grant from the National Natural Science Foundation of China (No.: 81271653).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Li, J., Zhang, R., Shi, L., Wang, D. (2017). Automatic Whole-Heart Segmentation in Congenital Heart Disease Using Deeply-Supervised 3D FCN. In: Zuluaga, M., Bhatia, K., Kainz, B., Moghari, M., Pace, D. (eds) Reconstruction, Segmentation, and Analysis of Medical Images. RAMBO HVSMR 2016 2016. Lecture Notes in Computer Science(), vol 10129. Springer, Cham. https://doi.org/10.1007/978-3-319-52280-7_11
Download citation
DOI: https://doi.org/10.1007/978-3-319-52280-7_11
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-52279-1
Online ISBN: 978-3-319-52280-7
eBook Packages: Computer ScienceComputer Science (R0)