Rotation Equivariant CNNs for Digital Pathology

  • Bastiaan S. VeelingEmail author
  • Jasper Linmans
  • Jim Winkens
  • Taco Cohen
  • Max Welling
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11071)


We propose a new model for digital pathology segmentation, based on the observation that histopathology images are inherently symmetric under rotation and reflection. Utilizing recent findings on rotation equivariant CNNs, the proposed model leverages these symmetries in a principled manner. We present a visual analysis showing improved stability on predictions, and demonstrate that exploiting rotation equivariance significantly improves tumor detection performance on a challenging lymph node metastases dataset. We further present a novel derived dataset to enable principled comparison of machine learning models, in combination with an initial benchmark. Through this dataset, the task of histopathology diagnosis becomes accessible as a challenging benchmark for fundamental machine learning research.



We thank Geert Litjens, Jakub Tomczak, Dimitrios Mavroeidis and the anonymous reviewers especially for their insightful comments. This research was supported by Philips Research, the SURFSara Lisa cluster and the NVIDIA GPU Grant.


  1. 1.
    Liu, Y., et al.: Detecting cancer metastases on gigapixel pathology images (2017)Google Scholar
  2. 2.
    Litjens, G., et al.: A survey on deep learning in medical image analysis (2017)CrossRefGoogle Scholar
  3. 3.
    Bejnordi, B.E.: Stain specific standardization of Whole-Slide histopathological images. IEEE Trans. Med. Imaging 35(2), 404–415 (2016)CrossRefGoogle Scholar
  4. 4.
    Zeiler, M.D., Fergus, R.: Visualizing and understanding convolutional networks. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014. LNCS, vol. 8689, pp. 818–833. Springer, Cham (2014). Scholar
  5. 5.
    Cohen, T.S., et al.: Group equivariant convolutional networks (2016)Google Scholar
  6. 6.
    Worrall, D.E., et al.: Harmonic networks: deep translation and rotation equivariance. In: Proceedings of IEEE CVPR, vol. 2 (2017).
  7. 7.
    Weiler, M., et al.: Learning steerable filters for rotation equivariant CNNs (2017)Google Scholar
  8. 8.
    Dumont, B., et al.: Robustness of Rotation-Equivariant networks to adversarial perturbations (2018)Google Scholar
  9. 9.
    Ehteshami, B.B.: Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. JAMA 318(22), 2199–2210 (2017)CrossRefGoogle Scholar
  10. 10.
    Cireşan, D.C.: Mitosis detection in breast cancer histology images with deep neural networks. MICCAI 16(Pt 2), 411–418 (2013)Google Scholar
  11. 11.
    Lenc, K., et al.: Understanding image representations by measuring their equivariance and equivalence. In: 2015 IEEE CVPR, pp. 991–999 (2015)Google Scholar
  12. 12.
    Huang, G., et al.: Densely connected convolutional networks (2016)Google Scholar
  13. 13.
    Ioffe, S., et al.: Batch normalization: accelerating deep network training by reducing internal covariate shift. In: ICML, pp. 448–456 (2015)Google Scholar
  14. 14.
    Spanhol, F.A.: A dataset for breast cancer histopathological image classification. IEEE Trans. Biomed. Eng. 63(7), 1455–1462 (2016)CrossRefGoogle Scholar
  15. 15.
    Song, Y., Chang, H., Huang, H., Cai, W.: Supervised intra-embedding of fisher vectors for histopathology image classification. In: Descoteaux, M., Maier-Hein, L., Franz, A., Jannin, P., Collins, D.L., Duchesne, S. (eds.) MICCAI 2017. LNCS, vol. 10435, pp. 99–106. Springer, Cham (2017). Scholar
  16. 16.
    Kingma, D.P., et al.: Adam: a method for stochastic optimization (2014)Google Scholar
  17. 17.
    Wang, D., et al.: Deep learning for identifying metastatic breast cancer (2016)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  • Bastiaan S. Veeling
    • 1
    Email author
  • Jasper Linmans
    • 1
  • Jim Winkens
    • 1
  • Taco Cohen
    • 1
  • Max Welling
    • 1
  1. 1.University of AmsterdamAmsterdamThe Netherlands

Personalised recommendations