Training Medical Image Analysis Systems like Radiologists

  • Gabriel MaicasEmail author
  • Andrew P. Bradley
  • Jacinto C. Nascimento
  • Ian Reid
  • Gustavo Carneiro
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11070)


The training of medical image analysis systems using machine learning approaches follows a common script: collect and annotate a large dataset, train the classifier on the training set, and test it on a hold-out test set. This process bears no direct resemblance with radiologist training, which is based on solving a series of tasks of increasing difficulty, where each task involves the use of significantly smaller datasets than those used in machine learning. In this paper, we propose a novel training approach inspired by how radiologists are trained. In particular, we explore the use of meta-training that models a classifier based on a series of tasks. Tasks are selected using teacher-student curriculum learning, where each task consists of simple classification problems containing small training sets. We hypothesize that our proposed meta-training approach can be used to pre-train medical image analysis models. This hypothesis is tested on the automatic breast screening classification from DCE-MRI trained with weakly labeled datasets. The classification performance achieved by our approach is shown to be the best in the field for that application, compared to state of art baseline approaches: DenseNet, multiple instance learning and multi-task learning.


Meta-learning Curriculum learning Multi-task training Breast image analysis Breast screening Magnetic resonance imaging 


  1. 1.
    The Royal Australian and New Zealand College of Radiologists: Training in Clinical Radiology (2009)Google Scholar
  2. 2.
    Wang, X., Peng, Y., et al.: Chestx-ray8: hospital-scale chest x-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. In: CVPR (2017)Google Scholar
  3. 3.
    Bar, Y., Diamant, I., et al.: Deep learning with non-medical training used for chest pathology identification. In: Medical Imaging: Computer-Aided Diagnosis (2015)Google Scholar
  4. 4.
    Finn, C., Abbeel, P., Levine, S.: Model-agnostic meta-learning for fast adaptation of deep networks. In: ICML (2017)Google Scholar
  5. 5.
    Matiisen, T., Oliver, A., Cohen, T., Schulman, J.: Teacher-student curriculum learning. arXiv preprint arXiv:1707.00183 (2017)
  6. 6.
    Huang, G., Liu, Z.: Densely connected convolutional networks. In: CVPR (2017)Google Scholar
  7. 7.
    Zhu, W., Lou, Q., Vang, Y.S., Xie, X.: Deep multi-instance networks with sparse label assignment for whole mammogram classification. In: Descoteaux, M., Maier-Hein, L., Franz, A., Jannin, P., Collins, D.L., Duchesne, S. (eds.) MICCAI 2017. LNCS, vol. 10435, pp. 603–611. Springer, Cham (2017). Scholar
  8. 8.
    Xue, W., Brahm, G.: Full left ventricle quantification via deep multitask relationships learning. Med. Image Anal. 43, 54–65 (2018)CrossRefGoogle Scholar
  9. 9.
    Smith, R.A., Andrews, K.S.: Cancer screening in the United States, 2017: a review of current American cancer society guidelines and current issues in cancer screening. CA Cancer J. Clin. 67, 100–121 (2017)CrossRefGoogle Scholar
  10. 10.
    Vreemann, S., Gubern-Merida, A.: The frequency of missed breast cancers in women participating in a high-risk MRI screening program. Breast Cancer Res. Treat. 169, 323–331 (2018)CrossRefGoogle Scholar
  11. 11.
    Gubern-Mérida, A., Martí, R.: Automated localization of breast cancer in DCE-MRI. Med. Image Anal. 20, 265–274 (2015)CrossRefGoogle Scholar
  12. 12.
    Dalmış, M.U., Vreemann, S.: Fully automated detection of breast cancer in screening MRI using convolutional neural networks. J. Med. Imaging 5, 014502 (2018)CrossRefGoogle Scholar
  13. 13.
    Amit, G., et al.: Hybrid mass detection in breast MRI combining unsupervised saliency analysis and deep learning. In: Descoteaux, M., Maier-Hein, L., Franz, A., Jannin, P., Collins, D.L., Duchesne, S. (eds.) MICCAI 2017. LNCS, vol. 10435, pp. 594–602. Springer, Cham (2017). Scholar
  14. 14.
    Jäger, P.F., et al.: Revealing hidden potentials of the q-Space signal in breast cancer. In: Descoteaux, M., et al. (eds.) MICCAI 2017. LNCS, vol. 10433, pp. 664–671. Springer, Cham (2017). Scholar
  15. 15.
    Maicas, G., Carneiro, G., Bradley, A.P., Nascimento, J.C., Reid, I.: Deep reinforcement learning for active breast lesion detection from DCE-MRI. In: Descoteaux, M., Maier-Hein, L., Franz, A., Jannin, P., Collins, D.L., Duchesne, S. (eds.) MICCAI 2017. LNCS, vol. 10435, pp. 665–673. Springer, Cham (2017). Scholar
  16. 16.
    Gutiérrez, B., Peter, L., Klein, T., Wachinger, C.: A multi-armed bandit to smartly select a training set from big medical data. In: Descoteaux, M., Maier-Hein, L., Franz, A., Jannin, P., Collins, D.L., Duchesne, S. (eds.) MICCAI 2017. LNCS, vol. 10435, pp. 38–45. Springer, Cham (2017). Scholar
  17. 17.
    Bengio, Y., Louradour, J., et al.: Curriculum learning. In: ICML (2009)Google Scholar
  18. 18.
    Sharma, S., Jha, A.K., Hedge, P., Ravindran, B.: Learning to multi-task by active sampling. In: ICLR (2018)Google Scholar
  19. 19.
    McClymont, D., Mehnert, A., et al.: Fully automatic lesion segmentation in breast MRI using mean-shift and graph-cuts on a region adjacency graph. In: JMRI (2014)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  • Gabriel Maicas
    • 1
    Email author
  • Andrew P. Bradley
    • 2
  • Jacinto C. Nascimento
    • 3
  • Ian Reid
    • 1
  • Gustavo Carneiro
    • 1
  1. 1.Australian Institute for Machine Learning, School of Computer ScienceThe University of AdelaideAdelaideAustralia
  2. 2.Science and Engineering FacultyQueensland University of TechnologyBrisbaneAustralia
  3. 3.Institute for Systems and RoboticsInstituto Superior TecnicoLisbonPortugal

Personalised recommendations