Semi-supervised Learning for Network-Based Cardiac MR Image Segmentation
- 8.8k Downloads
Training a fully convolutional network for pixel-wise (or voxel-wise) image segmentation normally requires a large number of training images with corresponding ground truth label maps. However, it is a challenge to obtain such a large training set in the medical imaging domain, where expert annotations are time-consuming and difficult to obtain. In this paper, we propose a semi-supervised learning approach, in which a segmentation network is trained from both labelled and unlabelled data. The network parameters and the segmentations for the unlabelled data are alternately updated. We evaluate the method for short-axis cardiac MR image segmentation and it has demonstrated a high performance, outperforming a baseline supervised method. The mean Dice overlap metric is 0.92 for the left ventricular cavity, 0.85 for the myocardium and 0.89 for the right ventricular cavity. It also outperforms a state-of-the-art multi-atlas segmentation method by a large margin and the speed is substantially faster.
This research has been conducted using the UK Biobank Resource under Application Number 18545. This work is supported by EPSRC programme Grant (EP/P001009/1). H.S. is supported by a Research Fellowship from the Uehara Memorial Foundation. P.M.M. gratefully acknowledges support from the Imperial College Healthcare Trust Biomedical Research Centre, the EPSRC Centre for Mathematics in Precision Healthcare and the MRC.
- 4.Chen, H., et al.: DCAN: deep contour-aware networks for accurate gland segmentation. In: CVPR, pp. 2487–2496 (2016)Google Scholar
- 5.Chen, L., et al.: DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. arXiv:1606.00915 (2016)
- 6.Krähenbühl, P., Koltun, V.: Efficient inference in fully connected CRFs with Gaussian edge potentials. In: NIPS, pp. 1–9 (2011)Google Scholar
- 7.Lin, D., et al.: ScribbleSup: scribble-supervised convolutional networks for semantic segmentation. In: CVPR, pp. 3159–3167 (2016)Google Scholar
- 8.Long, J., et al.: Fully convolutional networks for semantic segmentation. In: CVPR, pp. 3431–3440 (2015)Google Scholar
- 10.Papandreou, G., et al.: Weakly- and semi-supervised learning of a deep convolutional network for semantic image segmentation. In: ICCV, pp. 1742–1750 (2015)Google Scholar
- 11.Poudel, R., et al.: Recurrent fully convolutional neural networks for multi-slice MRI cardiac segmentation. arXiv:1608.03974 (2016)
- 14.Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale image recognition. In: ICLR, pp. 1–14 (2015)Google Scholar
- 15.Theano Development Team: Theano: a python framework for fast computation of mathematical expressions. arXiv:1605.02688 (2016)
- 16.Tran, P.: A fully convolutional neural network for cardiac segmentation in short-axis MRI. arXiv:1604.00494 (2016)
- 17.Yang, H., Sun, J., Li, H., Wang, L., Xu, Z.: Deep fusion net for multi-atlas segmentation: application to cardiac MR images. In: Ourselin, S., Joskowicz, L., Sabuncu, M.R., Unal, G., Wells, W. (eds.) MICCAI 2016. LNCS, vol. 9901, pp. 521–528. Springer, Cham (2016). doi: 10.1007/978-3-319-46723-8_60CrossRefGoogle Scholar