Skip to main content
SpringerLink
Log in

Improving multi-label chest X-ray disease diagnosis by exploiting disease and health labels dependencies

  • Published:
Multimedia Tools and Applications Aims and scope Submit manuscript

Abstract

The widely used ChestX-ray14 dataset addresses an important medical image classification problem and has the following caveats: 1) many lung pathologies are visually similar, 2) a variant of multiple diseases including lung cancer, tuberculosis, and pneumonia are present in a single scan at the same time, i.e. multiple labels. Existing literature uses state-of-the-art deep learning models being transfer learned where output neurons of the networks are trained for individual diseases to cater for multiple disease labels in each image. However, most of them don’t consider the label relationship explicitly between present and absent classes. In this work we have proposed a pair of novel error functions that can be employed for any deep learning model, Multi-label Softmax Loss (MSML) and Correlation Loss (CorLoss), to specifically address the properties of multiple labels and visually similar data. Moreover, we provide a fine-grained perspective into this problem and use bilinear pooling as an encoding scheme to increase discrimination of the model. The experiments are conducted on the ChestX-ray14 dataset. We first report improvements using our proposed loss with various backbone networks. After that, we extend our experiments to prove the rich disparity being learned by the model with our proposed losses, which can be fused with other models to improve the overall performances.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Notes

  1. We will omit i in the rest of the article for simplification.

  2. Overall performance of number-of-labels dependent results are lower than baseline because health images are removed. 5-Label, 6-Label and 7-Label subset is not reported due to small number of samples.

  3. We inherit the α = 0.1 and β = 0.3 from Section 3.4 and then equally weighted gradients are used for bilinear model training

  4. To facilitate the verification process, we use a small model ResNet18 for this experiment.

References

  1. Arandjelovic R, Gronat P, Torii A, Pajdla T, Sivic J (2016) Netvlad: Cnn architecture for weakly supervised place recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 5297–5307

  2. Armato SG, McLennan G, Bidaut L, McNitt-Gray MF, Meyer CR, Reeves AP, Zhao B, Aberle DR, Henschke CI, Hoffman EA et al (2011) The lung image database consortium (lidc) and image database resource initiative (idri): a completed reference database of lung nodules on ct scans. Medical physics 38(2):915–931

    Article  Google Scholar 

  3. Farrell R, Oza O, Zhang N, Morariu VI, Darrell T, Davis LS (2011) Birdlets: Subordinate categorization using volumetric primitives and pose-normalized appearance. In: in ICCV, IEEE, pp 161–168

  4. Ge ZY, McCool C, Sanderson C, Corke P (2015) Subset feature learning for fine-grained category classification. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, pp 46–52

  5. Ge ZY, McCool C, Sanderson C, Wang P, Liu L, Reid I, Corke P (2016) Exploiting temporal information for dcnn-based fine-grained object classification. arXiv:1608.00486

  6. Gong Y, Jia Y, Leung T, Toshev A, Ioffe S (2013) Deep convolutional ranking for multilabel image annotation. arXiv:1312.4894

  7. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: CVPR

  8. Huang G, Liu Z, Van Der Maaten L, Weinberger KQ (2017) Densely connected convolutional networks. In: CVPR

  9. Jaderberg M, Simonyan K, Zisserman A et al (2015) Spatial transformer networks. In: Advances in neural information processing systems, pp 2017–2025

  10. Kingma Diederik P, Ba Jimmy (2014) Adam: A method for stochastic optimization. arXiv:1412.6980

  11. Krause J, Jin H, Yang J, Li F-F (2015) Fine-grained recognition without part annotations. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 5546–5555

  12. Lin T-Y, RoyChowdhury A, Maji S (2015) Bilinear CNN models for fine-grained visual recognition. In: ICCV

  13. Lu H, Li Y, Chen M, Kim H, Serikawa S (2018) Brain intelligence: go beyond artificial intelligence. Mobile Networks and Applications 23(2):368–375

    Article  Google Scholar 

  14. Luke MJ, Mehrizi A, Folger GM, Rowe RD (1966) Chronic nasopharyngeal obstruction as a cause of cardiomegaly, cor pulmonale, and pulmonary edema. Pediatrics 37(5):762–768

    Google Scholar 

  15. Mahapatra D, Ge Z, Sedai S, Chakravorty R (2018) Joint registration and segmentation of xray images using generative adversarial networks. In: International Workshop on Machine Learning in Medical Imaging, Springer, pp 73–80

  16. Martins A, Astudillo R (2016) From softmax to sparsemax: A sparse model of attention and multi-label classification. In: International Conference on Machine Learning, pp 1614–1623

  17. Payer C, Štern D, Bischof H, Urschler M (2016) Regressing heatmaps for multiple landmark localization using cnns. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, pp 230–238

  18. Rajpurkar P, Irvin J, Zhu K, Yang B, Mehta H, Duan T, Ding D, Bagul A, langlotz C, Shpanskaya K et al (2017) Chexnet: Radiologist-level pneumonia detection on chest x-rays with deep learning. arXiv:1711.05225

  19. Rumelhart DE, Hinton GE, Williams RJ (1985) Learning internal representations by error propagation, Tech. Rep., California Univ San Diego La Jolla Inst for Cognitive Science

  20. Russakovsky O, Deng J, Su H, Krause J, Satheesh S, Ma S, Huang Z, Karpathy A, Khosla A, Bernstein M et al (2014) Imagenet large scale visual recognition challenge. Int J Comput Vis 115(3):211–252

    Article  MathSciNet  Google Scholar 

  21. Sa I, Ge Z, Dayoub F, Upcroft B, Perez T, McCool C (2016) Deepfruits: a fruit detection system using deep neural networks. Sensors 16(8):1222

    Article  Google Scholar 

  22. Sedai S, Mahapatra D, Ge Z, Chakravorty R, Garnavi R (2018) Deep multiscale convolutional feature learning for weakly supervised localization of chest pathologies in x-ray images. In: International Workshop on Machine Learning in Medical Imaging, Springer, pp 267–275

  23. Verbeek J Discriminative metric learning in nearest neighbor models for image annotation

  24. Wang X, Peng Y, Lu L, Lu Z, Bagheri M, Summers RM (2017) Chestx-ray8: Hospital-scale chest x-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. In: 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, pp 3462–3471

  25. Wei X-S, Luo J-H, Wu J, Zhou Z-H (2017) Selective convolutional descriptor aggregation for fine-grained image retrieval. IEEE Trans Image Process 26(6):2868–2881

    Article  MathSciNet  Google Scholar 

  26. Xu Z, Yang Y, Hauptmann AG (2015) A discriminative cnn video representation for event detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 1798–1807

  27. Xue X, Zhang W, Zhang J, Wu B, Fan J, Lu Y (2011) Correlative multi-label multi-instance image annotation. In: 2011 IEEE International Conference on Computer Vision (ICCV), IEEE, pp 651–658

  28. Yao L, Poblenz E, Dagunts D, Covington B, Bernard D, Lyman K (2017) Learning to diagnose from scratch by exploiting dependencies among labels. arXiv:1710.10501

  29. Zhang H, Xu T, Elhoseiny M, Huang X, Zhang S, Elgammal A, Metaxas D (2016) Spda-cnn: Unifying semantic part detection and abstraction for fine-grained recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 1143–1152

  30. Zhang N, Donahue J, Girshick R, Darrell T (2014) Part-based r-cnns for fine-grained category detection. In: ECCV

  31. Zhou Q, Yang W, Gao G, Ou W, Lu H, Chen J, Latecki LJ (2018) Multi-scale deep context convolutional neural networks for semantic segmentation. World Wide Web, pp 1–16

Download references

Acknowledgements

We would like to acknowledge the Airdoc for research funding support. The authors acknowledge Zitong Huang for driving useful discussions and support for the project. We also thank Nvidia AI Technology Centre for providing technical and hardware support for this project.

Author information

Authors and Affiliations

  1. Monash University, Clayton, Australia

    Zongyuan Ge & Xiaojun Chang

  2. Airdoc Research, Melbourne, Australia

    Zongyuan Ge

  3. Nividia AI Technology Centre, Edinburgh, UK

    Zongyuan Ge

  4. IBM Research, Melbourne, Australia

    Dwarikanath Mahapatra

  5. ETH Zurich, Zürich, Switzerland

    Zetao Chen

  6. La Trobe University, Bundoora, Australia

    Lianhua Chi

  7. Kyushu Institute Technology, Kitakyushu, Japan

    Huimin Lu

Authors
  1. Zongyuan Ge
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Dwarikanath Mahapatra
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Xiaojun Chang
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Zetao Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Lianhua Chi
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Huimin Lu
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Zongyuan Ge.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ge, Z., Mahapatra, D., Chang, X. et al. Improving multi-label chest X-ray disease diagnosis by exploiting disease and health labels dependencies. Multimed Tools Appl 79, 14889–14902 (2020). https://doi.org/10.1007/s11042-019-08260-2

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11042-019-08260-2

Keywords

Search

Navigation