Ensemble of Multi-sized FCNs to Improve White Matter Lesion Segmentation

  • Zhewei Wang
  • Charles D. Smith
  • Jundong LiuEmail author
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11046)


In this paper, we develop a two-stage neural network solution for the challenging task of white-matter lesion segmentation. To cope with the vast variability in lesion sizes, we sample brain MR scans with patches at three different dimensions and feed them into separate fully convolutional neural networks (FCNs). In the second stage, we process large and small lesion separately, and use ensemble-nets to combine the segmentation results generated from the FCNs. A novel activation function is adopted in the ensemble-nets to improve the segmentation accuracy measured by Dice Similarity Coefficient. Experiments on MICCAI 2017 White Matter Hyperintensities (WMH) Segmentation Challenge data demonstrate that our two-stage-multi-sized FCN approach, as well as the new activation function, are effective in capturing white-matter lesions in MR images.


  1. 1.
    Anbeek, P., et al.: Automatic segmentation of different-sized white matter lesions by voxel probability estimation. Med. Image Anal. 8(3), 205–215 (2004)CrossRefGoogle Scholar
  2. 2.
    Chen, Y., Shi, B., Wang, Z., Sun, T., Smith, C.D., Liu, J.: Accurate and Consistent Hippocampus Segmentation Through Convolutional LSTM and View Ensemble. In: Wang, Q., Shi, Y., Suk, H.-I., Suzuki, K. (eds.) MLMI 2017. LNCS, vol. 10541, pp. 88–96. Springer, Cham (2017). Scholar
  3. 3.
    Chen, Y., Shi, B., Wang, Z., Zhang, P., Smith, C.D., Liu, J.: Hippocampus segmentation through multi-view ensemble ConvNets. In: 14th IEEE, ISBI 2017, pp. 192–196 (2017)Google Scholar
  4. 4.
    Çiçek, Ö., et al.: 3D U-Net: learning dense volumetric segmentation from sparse annotation. In: MICCAI, pp. 424–432. Springer, Berlin (2016)CrossRefGoogle Scholar
  5. 5.
    Drozdzal, M., et al.: The importance of skip connections in biomedical image segmentation. In: Deep Learning and Data Labeling for Medical Applications, pp. 179–187. Springer, Chem (2016)Google Scholar
  6. 6.
    Ekanayake, J., et al.: Generalised wasserstein dice score for imbalanced multi-class segmentation using holistic convolutional networks. In: Third International Workshop, BrainLes, MICCAI (2017)Google Scholar
  7. 7.
    Li, H., et al.: Fully convolutional network ensembles for white matter hyperintensities segmentation in MR images. arXiv preprint arXiv:1802.05203 (2018)
  8. 8.
    Liu, J., Smith, C.D., Chebrolu, H.: Automatic multiple sclerosis detection based on integrated square estimation. In: IEEE, CVPR Workshops, 20–25 June, 2009, pp. 31–38 (2009)Google Scholar
  9. 9.
    Long, J., et al.: Fully convolutional networks for semantic segmentation. In: Proceedings of CVPR, pp. 3431–3440 (2015)Google Scholar
  10. 10.
    Milletari, F., et al.: V-net: fully convolutional neural networks for volumetric medical image segmentation. In: 3D Vision (3DV). pp. 565–571. IEEE (2016)Google Scholar
  11. 11.
    Ronneberger, O., Fischer, P., Brox, T.: U-net: convolutional networks for biomedical image segmentation. In: MICCAI, pp. 234–241. Springer, Berlin (2015)Google Scholar
  12. 12.
    Souplet, J.C., et al.: An automatic segmentation of T2-FLAIR multiple sclerosis lesions. In: The MIDAS Journal-MS Lesion Segmentation (MICCAI 2008 Workshop). Citeseer (2008)Google Scholar
  13. 13.
    Sudre, C.H., et al.: Generalised dice overlap as a deep learning loss function for highly unbalanced segmentations. In: Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support, pp. 240–248. Springer, Berlin (2017)CrossRefGoogle Scholar
  14. 14.
    Warfield, S., Tomas-Fernandez, X.: Lesion segmentation. In: Toga, A.W. (ed.) Brain Mapping, pp. 323–332. Academic Press, Waltham (2015)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  1. 1.School of Electrical Engineering and Computer ScienceOhio UniversityAthensUSA
  2. 2.Department of NeurologyUniversity of KentuckyLexingtonUSA

Personalised recommendations