Automated lung cancer diagnosis using three-dimensional convolutional neural networks

Abstract

Lung cancer is the deadliest cancer worldwide. It has been shown that early detection using low-dose computer tomography (LDCT) scans can reduce deaths caused by this disease. We present a general framework for the detection of lung cancer in chest LDCT images. Our method consists of a nodule detector trained on the LIDC-IDRI dataset followed by a cancer predictor trained on the Kaggle DSB 2017 dataset and evaluated on the IEEE International Symposium on Biomedical Imaging (ISBI) 2018 Lung Nodule Malignancy Prediction test set. Our candidate extraction approach is effective to produce accurate candidates with a recall of 99.6%. In addition, our false positive reduction stage classifies successfully the candidates and increases precision by a factor of 2000. Our cancer predictor obtained a ROC AUC of 0.913 and was ranked 1st place at the ISBI 2018 Lung Nodule Malignancy Prediction challenge.

Graphical abstract

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14

Notes

  1. 1.

    LIDC-IDRI can be found at The Cancer Imaging Archive (TCIA): https://wiki.cancerimagingarchive.net/display/Public/LIDC-IDRI

  2. 2.

    NLST can be found at The Cancer Imaging Archive (TCIA): https://wiki.cancerimagingarchive.net/display/NLST

  3. 3.

    Method by Mehrtash et al. Currently not published. Information can be found at: https://www.rsipvision.com/ComputerVisionNews-2018May/28/

  4. 4.

    ISBI 2018 lung cancer challenge results can be found at: https://bit.ly/2JPNnGS

  5. 5.

    https://github.com/BCV-Uniandes/LungCancerDiagnosis-pytorch

References

  1. 1.

    Aberle D et al (2011) Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 365:395–409. https://doi.org/10.1056/NEJMoa1102873

    Article  PubMed  Google Scholar 

  2. 2.

    American Cancer Society (2015) American Cancer Society. Cancer Facts and Figures 2015. American Cancer Society, Atlanta

    Google Scholar 

  3. 3.

    Armato S III 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. Med Phys 38:915–931

    Article  Google Scholar 

  4. 4.

    Chen S, et al., 2016. Automatic scoring of multiple semantic attributes with multi-task feature leverage: a study on pulmonary nodules in CT images. IEEE Trans Med Imaging, in press.

  5. 5.

    Ciompi F et al (2015) Automatic classification of pulmonary peri-fissural nodules in computed tomography using an ensemble of 2D views and a convolutional neural network out-of-the-box. Med Image Anal 26:195–202

    Article  Google Scholar 

  6. 6.

    Ciompi F, et al., 2016. Towards automatic pulmonary nodule management in lung cancer screening with deep learning. arXiv:1610.09157.

  7. 7.

    Ciompi F, et al., 2017. Towards automatic pulmonary nodule management in lung cancer screening with deep learning.

  8. 8.

    Dalal N, Triggs B, 2005. Histograms of oriented gradients for human detection, in: In CVPR, pp. 886–893.

  9. 9.

    Dolejsi M, Kybic J, 2007. Automatic two-step detection of pulmonary nodules. Proceedings of SPIE 6514, 3j–1–3j–12.

  10. 10.

    Dou Q, et al., 2016. Multi-level contextual 3D CNNs for false positive reduction in pulmonary nodule detection, in press.

  11. 11.

    Falk S, Williams C, 2010. “Chapter 1”. Lung cancer—the facts (3rd ed.). Oxford University Press. pp.3–4.

  12. 12.

    He K, et al., 2015. Deep residual learning for image recognition. CoRR abs/1512.03385. ArXiv:1512.03385.

  13. 13.

    Hu Z et al (2018) Deep learning for image-based cancer detection and diagnosis–a survey. Pattern Recognition Volume 83:134–149

    Article  Google Scholar 

  14. 14.

    Hussein S, et al., 2017. Risk stratification of lung nodules using 3D CNN-based multi-task learning. IPMI .

  15. 15.

    Jacobs C et al (2014a) Automatic detection of subsolid pulmonary nodules in thoracic computed tomography images. Med Image Anal 18:374–384

    Article  Google Scholar 

  16. 16.

    Kaggle Data Science Bowl, 2017. Data Science Bowl 2017. https://www.kaggle.com/c/data-science-bowl-2017.

  17. 17.

    Krizhevsky A, et al., 2012. Imagenet classification with deep convolutional neural networks, in: Advances in Neural Information Processing Systems 25. Curran Associates, Inc., pp. 1097–1105.

  18. 18.

    Lee S et al (2012) Automated detection of ling nodules in computed tomography images: a review. Mach Vis Appl 23:151–163

    Article  Google Scholar 

  19. 19.

    Li Q (2007) Recent progress in computer-aided diagnosis of lung nodules on thin-section CT. Comput Med Imaging Graph 31:248–257

    Article  Google Scholar 

  20. 20.

    Li W, et al., 2016. Pulmonary nodule classification with deep convolutional neural networks on computed tomography images. Computational and Mathematical Methods in Medicine, 6215085.

  21. 21.

    Litjents G et al (2017) A survey on deep learning in medical image analysis. Med Image Anal 42:60–88

    Article  Google Scholar 

  22. 22.

    Lung Nodule Malignancy Prediction Challenge, 2018. ISBI 2018 - lung nodule malignancy prediction, based on sequential CT scans. http://isbichallenges.cloudapp.net/competitions.

  23. 23.

    Massion PP, Walker RC (2014) Indeterminate pulmonary nodules: risk for having or for developing lung cancer? Cancer Prev Res (Phila) 7(12):1173–1178. https://doi.org/10.1158/1940-6207.CAPR-14-0364

    Article  Google Scholar 

  24. 24.

    Meyer P et al (2018 Jul 1) 2017. Survey on deep learning for radiotherapy. Comput Biol Med 98:126–146. https://doi.org/10.1016/j.compbio-med.2018.05.018

    Article  PubMed  Google Scholar 

  25. 25.

    Moria S et al (2018) Blood vessel segmentation algorithms — review of methods, datasets and evaluation metrics. Comput Methods Prog Biomed. https://doi.org/10.1016/j.cmpb.2018.02.001

  26. 26.

    Murphy K et al (2009) A large scale evaluation of automatic pulmonary nodule detection in chest CT using local image features and k-nearest-neighbour classification. Med Image Anal 13:757–770

    CAS  Article  Google Scholar 

  27. 27.

    NCI. 12 May 2015. Non-small cell lung cancer treatment –patient version. Archived from the original on 29 February 2016. Retrieved 5 March 2016.

  28. 28.

    Nair V, Hinton G, 2010. Rectified linear units improve restricted Boltzmann machines, in: Proceedings of the 27th International Conference on International Conference on Machine Learning, Omnipress, USA. pp. 807–814.

  29. 29.

    Oseas A et al (2014) Automatic detection of solitary lung nodules using quality threshold clustering, genetic algorithm and diversity index. Artif Intell Med 60:165–177

    Article  Google Scholar 

  30. 30.

    Osman S, Ucan O (2007) Lung nodule diagnosis using 3d template matching. Comput Biol Med 37:1167–1172

    Article  Google Scholar 

  31. 31.

    Pastorino U, et al., 2012. Annual or biennal CT screening versus observation in heavy smokers: 5-year results of the MILD trial. European Journal of Cancer Prevention.

  32. 32.

    Perez G, Arbelaez P, 2017. Automated detection of lung nodules with three-dimensional convolutional neural networks. Doi:https://doi.org/10.1117/12.2285954.

  33. 33.

    Ren S, et al., 2015. Faster r-cnn: Towards real-time object detection with region proposal networks, in: Advances in Neural Information Processing Systems 28. Curran Associates, Inc., pp. 91–99.

  34. 34.

    Setio A et al (2015) Automatic detection of large pulmonary solid nodules in thoracic CT images. Med Phys 42:5642–5653

    Article  Google Scholar 

  35. 35.

    Setio A et al (2016) Pulmonary nodule detection in CT images: false positive reduction using multi-view convolutional networks. IEEE Trans Med Imaging 35(5):1160–1169

    Article  Google Scholar 

  36. 36.

    Shelhamer E et al (2017) Fully convolutional networks for semantic segmentation. IEEE Trans Pattern Anal Mach Intell 39:640–651. https://doi.org/10.1109/TPAMI.2016.2572683

    Article  PubMed  Google Scholar 

  37. 37.

    Shen W, et al., 2015. Multi-scale convolutional neural networks for lung nodule classification. In: Inf Process Med Imaging. Vol. 9123 of Lect Notes Comput Sci. pp. 588–599.

  38. 38.

    Shen W, et al., 2016. Learning from experts: developing transferable deep features for patient-level lung cancer prediction. In: Med Image Comput Comput Assist Interv. Vol. 9901 of Lect Notes Comput Sci. pp. 124–131.

  39. 39.

    Simonyan K, Zisserman A, 2014. Very deep convolutional networks for large-scale image recognition. CoRR abs/1409.1556. ArXiv:1409.1556.

  40. 40.

    Song K, et al., 2015. Using deep learning for classification of lung nodules on computed tomography images.

  41. 41.

    Sudha V, Jayashree P (2012) Lung nodule detection in CT images using thresholding and morphological operations. International Journal on Emerging Science and Engineering (IJESE) 1:17–21

    Google Scholar 

  42. 42.

    Sun W, et al., 2016. Computer aided lung cancer diagnosis with deep learning algorithms. Research gate.

  43. 43.

    Szegedy C, et al., 2015. Going deeper with convolutions, in: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

    Google Scholar 

  44. 44.

    Tan M et al (2011) A novel computer-aided lung nodule detection system for CT images. Med Phys 38:5630–5645

    Article  Google Scholar 

  45. 45.

    Tartar A, Akan A, 2013. A new method for pulmonary nodule detection using decision trees. 35th Annual International Conference of the IEEE EMBS .

  46. 46.

    Teramoto A et al (2016) Automated detection of pulmonary nodules in PET/CT images: ensemble false-positive reduction using a convolutional neural network technique. Med Phys 43:2821–2827

    Article  Google Scholar 

  47. 47.

    Torres E et al (2015) Large scale validation of the M5L lung CAD on heterogeneous CT datasets. Med Phys 42:1477–1489

    Article  Google Scholar 

  48. 48.

    van Ginneken B et al (2010) Comparing and combining algorithms for computer-aided detection of pulmonary nodules in computed tomography scans: the ANODE09 study. Med Image Anal 14:707–722

    Article  Google Scholar 

  49. 49.

    van Ginneken B, et al., 2015. Off-the shelf convolutional neural network features for pulmonary nodule detection in computed tomography scans. In: IEEE Int Symp Biomedical Imaging. pp. 286–289.

  50. 50.

    World Health Organization (2017) World Health Organization. Media Centre, Cancer http://www.who.int/mediacentre/factsheets/fs297/en/

    Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Gustavo Perez.

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

Verify currency and authenticity via CrossMark

Cite this article

Perez, G., Arbelaez, P. Automated lung cancer diagnosis using three-dimensional convolutional neural networks. Med Biol Eng Comput 58, 1803–1815 (2020). https://doi.org/10.1007/s11517-020-02197-7

Download citation

Keywords

  • Computed tomography
  • Computer-aided diagnosis
  • Convolutional neural networks
  • Deep learning
  • Lung cancer