Abstract
Automated pathology segmentation remains a valuable diagnostic tool in clinical practice. However, collecting training data is challenging. Semi-supervised approaches by combining labelled and unlabelled data can offer a solution to data scarcity. An approach to semi-supervised learning relies on reconstruction objectives (as self-supervision objectives) that learns in a joint fashion suitable representations for the task. Here, we propose Anatomy-Pathology Disentanglement Network (APD-Net), a pathology segmentation model that attempts to learn jointly for the first time: disentanglement of anatomy, modality, and pathology. The model is trained in a semi-supervised fashion with new reconstruction losses directly aiming to improve pathology segmentation with limited annotations. In addition, a joint optimization strategy is proposed to fully take advantage of the available annotations. We evaluate our methods with two private cardiac infarction segmentation datasets with LGE-MRI scans. APD-Net can perform pathology segmentation with few annotations, maintain performance with different amounts of supervision, and outperform related deep learning methods.
H. Jiang — Most of the work was completed in the School of Engineering at the University of Edinburgh
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.
We only consider unlabeled pathology, assuming anatomy masks are available during training. Partial anatomy annotation is out of the scope of this paper.
- 2.
Note that \(p^i\) is the same as \(\hat{y}^i_{pat}\), i.e. the predicted pathology mask. We use \(p^i\) for disentanglement and image reconstruction, and \(\hat{y}^i_{pat}\) for pathology segmentation.
- 3.
The Dice loss can be seen as a special case of the Tversky loss [15] when \(\beta =0.5\).
- 4.
The \(\mathbb {1}\) matrix is added to ensure that no zero elements are multiplied with \(\Vert x^i-\hat{x}^i\Vert _1\). Also, if \(\lambda _{pat}=1\), the loss reduces to the \(\ell _1\) loss.
- 5.
U-Net (masked / unmasked) and Cascaded U-Net are optimized with full supervision using Tversky and focal losses, and penalized as defined in the Training details. In reality, U-Net (masked) is not a good choice since manual myocardial annotations are not always available at inference time.
- 6.
Code will be available at https://github.com/falconjhc/APD-Net shortly.
References
Abraham, N., Khan, N.M.: A novel focal Tversky loss function with improved attention U-Net for lesion segmentation. In: 2019 IEEE 16th International Symposium on Biomedical Imaging. (ISBI 2019), pp. 683–687. IEEE (2019)
Chartsias, A., et al.: Factorised spatial representation learning: application in semi-supervised Myocardial segmentation. In: Frangi, A.F., Schnabel, J.A., Davatzikos, C., Alberola-López, C., Fichtinger, G. (eds.) MICCAI 2018. LNCS, vol. 11071, pp. 490–498. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-00934-2_55
Chartsias, A., et al.: Disentangled representation learning in cardiac image analysis. Med. Image Anal. 58, 101535 (2019)
Chartsias, A., et al.: Disentangle, align and fuse for multimodal and zero-shot image segmentation. arXiv preprint arXiv:1911.04417 (2019)
Dey, R., Hong, Y.: CompNet: complementary segmentation network for brain MRI extraction. In: Frangi, A.F., Schnabel, J.A., Davatzikos, C., Alberola-López, C., Fichtinger, G. (eds.) MICCAI 2018. LNCS, vol. 11072, pp. 628–636. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-00931-1_72
Huang, X., Liu, M.Y., Belongie, S., Kautz, J.: Multimodal unsupervised image-to-image translation. In: Proceedings of the European Conference on Computer Vision (ECCV), pp. 172–189 (2018)
Karim, R., et al.: Evaluation of state-of-the-art segmentation algorithms for left ventricle infarct from late Gadolinium enhancement MR images. Med. Image Anal. 30, 95–107 (2016)
Lin, T.Y., Goyal, P., Girshick, R., He, K., Dollár, P.: Focal loss for dense object detection. In: Proceedings of the IEEE international conference on computer vision, pp. 2980–2988 (2017)
Moccia, S., et al.: Development and testing of a deep learning-based strategy for scar segmentation on CMR-LGE images. Magn. Reson. Mater. Phys. Biol. Med. 32(2), 187–195 (2018). https://doi.org/10.1007/s10334-018-0718-4
Morshid, A., et al.: A machine learning model to predict hepatocellular carcinoma response to transcatheter arterial chemoembolization. Radiol. Artif. Intell. 1(5), e180021 (2019)
Pang, Y., Hu, D., Sun, M.: A modified scheme for liver tumor segmentation based on cascaded FCNs. In: Proceedings of the International Conference on Artificial Intelligence, Information Processing and Cloud Computing, pp. 1–6 (2019)
Qin, C., Shi, B., Liao, R., Mansi, T., Rueckert, D., Kamen, A.: Unsupervised deformable registration for multi-modal images via disentangled representations. In: Chung, A.C.S., Gee, J.C., Yushkevich, P.A., Bao, S. (eds.) IPMI 2019. LNCS, vol. 11492, pp. 249–261. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-20351-1_19
Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 234–241. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28
Roth, K., Konopczyński, T., Hesser, J.: Liver lesion segmentation with slice-wise 2D Tiramisu and Tversky loss function. arXiv preprint arXiv:1905.03639 (2019)
Salehi, S.S.M., Erdogmus, D., Gholipour, A.: Tversky loss function for image segmentation using 3D fully convolutional deep networks. In: Wang, Q., Shi, Y., Suk, H.-I., Suzuki, K. (eds.) MLMI 2017. LNCS, vol. 10541, pp. 379–387. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-67389-9_44
Schroff, F., Kalenichenko, D., Philbin, J.: FaceNet: a unified embedding for face recognition and clustering. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 815–823 (2015)
Tran, G.S., Nghiem, T.P., Nguyen, V.T., Luong, C.M., Burie, J.C.: Improving accuracy of lung nodule classification using deep learning with focal loss. J. Healthc. Eng. 2019 (2019)
van Tulder, G., de Bruijne, M.: Learning cross-modality representations from multi-modal images. IEEE Trans. Med. Imaging 38(2), 638–648 (2018)
Williams, R.J., Zipser, D.: A learning algorithm for continually running fully recurrent neural networks. Neural Comput. 1(2), 270–280 (1989)
Xia, T., Chartsias, A., Tsaftaris, S.A.: Pseudo-healthy synthesis with pathology disentanglement and adversarial learning. Med. Image Anal. 64, 101719 (2020)
Zakharov, S., Kehl, W., Planche, B., Hutter, A., Ilic, S.: 3D object instance recognition and pose estimation using triplet loss with dynamic margin. In: 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 552–559. IEEE (2017)
Acknowledgement
This work was supported by US National Institutes of Health (1R01HL136578-01). This work used resources provided by the Edinburgh Compute and Data Facility (http://www.ecdf.ed.ac.uk/). S.A. Tsaftaris acknowledges the Royal Academy of Engineering and the Research Chairs and Senior Research Fellowships scheme.
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
Jiang, H. et al. (2020). Semi-supervised Pathology Segmentation with Disentangled Representations. In: Albarqouni, S., et al. Domain Adaptation and Representation Transfer, and Distributed and Collaborative Learning. DART DCL 2020 2020. Lecture Notes in Computer Science(), vol 12444. Springer, Cham. https://doi.org/10.1007/978-3-030-60548-3_7
Download citation
DOI: https://doi.org/10.1007/978-3-030-60548-3_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-60547-6
Online ISBN: 978-3-030-60548-3
eBook Packages: Computer ScienceComputer Science (R0)
