Abstract
Myocardial pathology segmentation in cardiac magnetic resonance (CMR) is significant in the diagnosis for patients suffering from myocardial infarction (MI). Therefore, accurate and automatic segmentation method is highly desired in clinical practice. To better accomplish this segmentation task, we propose a modified U-net architecture named Dual Attention U-net. In this network, we use U-net as the baseline and embed a dual-branch attention module in it. One of the branches provides channel attention via emphasizing feature association among different channel maps, while the other branch provides spatial attention which adaptively aggregates the features at relative positions regardless of their distances in a weighted manner. Experiments show that both of these modules have effectively improved the segmentation performance. In addition, we have adopted data processing and augmentation methods to further improve the segmentation quality. Our model is evaluated on the public dataset from the MyoPS 2020 challenge, which consists of three sequences of cardiac MR images (bSSFP, LGE, and T2-weighted) from 45 patients. Our method achieves the Dice score of 63.5 (scar) and 68.8 (scar and edema) in the final test set.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Campello, V.M., Martinisla, C., Izquierdo, C., Petersen, S.E., Ballester, M.A.G., Lekadir, K.: Combining multi-sequence and synthetic images for improved segmentation of late gadolinium enhancement cardiac MRI. arXiv-Image and Video Processing (2019)
Chen, J., et al.: Multiview two-task recursive attention model for left atrium and atrial scars segmentation, pp. 455–463 (2018)
Fu, J., et al.: Dual attention network for scene segmentation, pp. 3146–3154 (2019)
Han, W.K., Farzaneh-Far, A., Kim, R.J.: Cardiovascular magnetic resonance in patients with myocardial infarction: current and emerging applications. J. Am. Coll. Cardiol. 55(1), 1–16 (2009)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition, pp. 770–778 (2016)
Khened, M., Alex, V., Krishnamurthi, G.: Fully convolutional multi-scale residual DenseNets for cardiac segmentation and automated cardiac diagnosis using ensemble of classifiers. Med. Image Anal. 51, 21–45 (2018)
Liu, Y., Wang, W., Wang, K., Ye, C., Luo, G.: An automatic cardiac segmentation framework based on multi-sequence MR image. arXiv-Image and Video Processing (2019)
Long, J., Shelhamer, E., Darrell, T.: Fully convolutional networks for semantic segmentation, pp. 3431–3440 (2015)
Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation. CoRR abs/1505.04597 (2015). http://arxiv.org/abs/1505.04597
Shi, Z., et al.: Bayesian VoxDRN: a probabilistic deep Voxelwise dilated residual network for whole heart segmentation from 3D MR images, pp. 569–577 (2018)
Wang, X., Girshick, R., Gupta, A., He, K.: Non-local neural networks, pp. 7794–7803 (2018)
Wang, X., et al.: SK-Unet: an improved u-net model with selective kernel for the segmentation of multi-sequence cardiac MR. arXiv-Image and Video Processing (2020)
Xiong, Z., Fedorov, V.V., Fu, X., Cheng, E., Macleod, R.S., Zhao, J.: Fully automatic left atrium segmentation from late gadolinium enhanced magnetic resonance imaging using a dual fully convolutional neural network. IEEE Trans. Med. Imaging 38(2), 515–524 (2019)
Zabihollahy, F., White, J.A., Ukwatta, E.: Myocardial scar segmentation from magnetic resonance images using convolutional neural network. In: Computer-Aided Diagnosis (2018)
Zhuang, X.: Multivariate mixture model for myocardial segmentation combining multi-source images. IEEE Trans. Pattern Anal. Mach. Intell. 41(12), 2933–2946 (2019)
Zhuang, X.: Multivariate mixture model for cardiac segmentation from multi-sequence MRI. In: International Conference on Medical Image Computing & Computer-assisted Intervention (2016)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Yu, H., Zha, S., Huangfu, Y., Chen, C., Ding, M., Li, J. (2020). Dual Attention U-Net for Multi-sequence Cardiac MR Images Segmentation. In: Zhuang, X., Li, L. (eds) Myocardial Pathology Segmentation Combining Multi-Sequence Cardiac Magnetic Resonance Images. MyoPS 2020. Lecture Notes in Computer Science(), vol 12554. Springer, Cham. https://doi.org/10.1007/978-3-030-65651-5_11
Download citation
DOI: https://doi.org/10.1007/978-3-030-65651-5_11
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-65650-8
Online ISBN: 978-3-030-65651-5
eBook Packages: Computer ScienceComputer Science (R0)