Skip to content
naulon
Get started

No retraction, correction or concern found when last checked on Oct 17, 2026.

Automatic lung segmentation in routine imaging is primarily a data diversity problem, not a methodology problem

Johannes Hofmanninger, Forian Prayer, Jeanny Pan and 3 more

European Radiology Experimental · 2020 · CC BY

doi:10.1186/s41747-020-00173-2

Abstract

Background. Automated segmentation of anatomical structures is a crucial step in image analysis. For lung segmentation in computed tomography, a variety of approaches exists, involving sophisticated pipelines trained and validated on different datasets. However, the clinical applicability of these approaches across diseases remains limited.

Methods. We compared four generic deep learning approaches trained on various datasets and two readily available lung segmentation algorithms. We performed evaluation on routine imaging data with more than six different disease patterns and three published data sets.

Results. Using different deep learning approaches, mean Dice similarity coefficients (DSCs) on test datasets varied not over 0.02. When trained on a diverse routine dataset (n = 36), a standard approach (U-net) yields a higher DSC (0.97 ± 0.05) compared to training on public datasets such as the Lung Tissue Research Consortium (0.94 ± 0.13, p = 0.024) or Anatomy 3 (0.92 ± 0.15, p = 0.001). Trained on routine data (n = 231) covering multiple diseases, U-net compared to reference methods yields a DSC of 0.98 ± 0.03 versus 0.94 ± 0.12 (p = 0.024).

Conclusions. The accuracy and reliability of lung segmentation algorithms on demanding cases primarily relies on the diversity of the training data, highlighting the importance of data diversity compared to model choice. Efforts in developing new datasets and providing trained models to the public are critical. By releasing the trained model under General Public License 3.0, we aim to foster research on lung diseases by providing a readily available tool for segmentation of pathological lungs.

Full text

Its licence does not allow us to show the full text. Read it at PubMed Central:

Read at PubMed Central

References 41

  1. 1.

    OECD (2017) Health at a Glance 2017: OECD indicators. 10.1787/19991312

    Not in our library

  2. 2.

    Mansoor A, Bagci U, Foster B et al (2015) Segmentation and image analysis of abnormal lungs at CT: current approaches, challenges, and future trends. Radiographics 35:1056–1076. 10.1148/rg.2015140232

    Not in our library10.1148/rg.2015140232

  3. 3.
    Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: A cross-sectional study · 2018
    No notice
  4. 4.

    Göksel O, Jiménez-del Toro OA, Foncubierta-Rodríguez A, Muller H (2015) Overview of the VISCERAL Challenge at ISBI. In: Proceedings of the VISCERAL Challenge at ISBI 2015. New York, NY

    Not in our library

  5. 5.

    Yang J, Veeraraghavan H, Armato SG 3rd et al (2018) Autosegmentation for thoracic radiation treatment planning: a grand challenge at AAPM 2017. Med Phys 45:4568–4581. 10.1002/mp.13141

    Not in our library10.1002/mp.13141

  6. 6.

    Oakden-Rayner L, Bessen T, Palmer LJ, Carneiro G, Nascimento JC, Bradley AP (2017) Precision radiology: predicting longevity using feature engineering and deep learning methods in a radiomics framework. Sci Rep 7. 10.1038/s41598-017-01931-w

    Not in our library10.1038/s41598-017-01931-w

  7. 7.

    Stein JM, Walkup LL, Brody AS, Fleck RJ, Woods JC. Quantitative CT characterization of pediatric lung development using routine clinical imaging. Pediatr Radiol. 2016;46:1804-1812.

    Not in our library10.1007/s00247-016-3686-8

  8. 8.

    Korfiatis P, Skiadopoulos S, Sakellaropoulos P, Kalogeropoulou C, Costaridou L. Combining 2D wavelet edge highlighting and 3D thresholding for lung segmentation in thin-slice CT. Br J Radiol. 2007;80:996-1004.

    Not in our library10.1259/bjr/20861881

  9. 9.

    Hu S, Hoffman EA, Reinhardt JM. Automatic lung segmentation for accurate quantitation of volumetric X-ray CT images. IEEE Trans Med Imaging. 2001;20:490-498.

    Not in our library10.1109/42.929615

  10. 10.

    Chen H, Mukundan R, Butler A (2011) Automatic lung segmentation in HRCT images. International Conference on Image and Vision Computing, In, pp 293–298 http://hdl.handle.net/10092/6246

    Not in our library

  11. 11.

    Pulagam AR, Kande GB, Ede VKR, Inampudi RB. Automated lung segmentation from HRCT scans with diffuse parenchymal lung diseases. J Digit Imaging. 2016;29:507-519.

    Not in our library10.1007/s10278-016-9875-z

  12. 12.

    Sluimer I, Prokop M, van Ginneken B. Toward automated segmentation of the pathological lung in CT. IEEE Trans Med Imaging. 2005;24:1025-1038.

    Not in our library10.1109/tmi.2005.851757

  13. 13.

    Iglesias JE, Sabuncu MR. Multi-atlas segmentation of biomedical images: a survey. Med Image Anal. 2015;24:205-219.

    Not in our library10.1016/j.media.2015.06.012

  14. 14.

    Li Z, Hoffman EA, Reinhardt JM. Atlas-driven lung lobe segmentation in volumetric X-ray CT images. IEEE Trans Med Imaging. 2005;25:1-16.

    Not in our library10.1109/tmi.2005.859209

  15. 15.

    Sun S, Bauer C, Beichel R. Automated 3-D segmentation of lungs with lung cancer in CT data using a novel robust active shape model approach. IEEE Trans Med Imaging. 2012;31:449-460.

    Not in our library10.1109/tmi.2011.2171357

  16. 16.

    Agarwala S, Nandi D, Kumar A, Dhara AK, Sadhu SBTA, Bhadra AK. Automated segmentation of lung field in HRCT images using active shape model. IEEE Region 10 Annual International Conference. 2017;2516-2520.

    Not in our library

  17. 17.

    Chen G, Xiang D, Zhang B et al (2019) Automatic pathological lung segmentation in low-dose CT image using eigenspace sparse shape composition. IEEE Trans Med Imaging 38:1736–1749. 10.1109/TMI.2018.2890510

    Not in our library10.1109/tmi.2018.2890510

  18. 18.

    Sofka M, Wetzl J, Birkbeck N et al (2011) Multi-stage learning for robust lung segmentation in challenging CT volumes. In: Fichtinger G, Martel A, Peters T (eds) International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, Berlin, Heidelberg, pp 667–674. 10.1007/978-3-642-23626-6_82

    Not in our library10.1007/978-3-642-23626-6_82

  19. 19.

    Harrison AP, Xu Z, George K, Lu L, Summers RM, Mollura DJ (2017) Progressive and multi-path holistically nested neural networks for pathological lung segmentation from CT images. In: Descoteaux M, Maier-Hein L, Franz A, Jannin P, Collins D, Duchesne S (eds) International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, Cham, pp 621–629. 10.1007/978-3-319-66179-7_71

    Not in our library

  20. 20.

    Korfiatis P, Kalogeropoulou C, Karahaliou A, Kazantzi A, Skiadopoulos S, Costaridou L. Texture classification-based segmentation of lung affected by interstitial pneumonia in high-resolution CT. Med Phys. 2008;35:5290-5302.

    Not in our library10.1118/1.3003066

  21. 21.

    Wang J, Li F, Li Q. Automated segmentation of lungs with severe interstitial lung disease in CT. Med Phys. 2009;36:4592-4599.

    Not in our library10.1118/1.3222872

  22. 22.

    Soliman A, Khalifa F, Elnakib A et al (2017) Accurate lungs segmentation on CT chest images by adaptive appearance-guided shape modeling. IEEE Trans Med Imaging 36:263–276. 10.1109/TMI.2016.2606370

    Not in our library10.1109/tmi.2016.2606370

  23. 23.

    van Rikxoort EM, de Hoop B, Viergever MA, Prokop M, van Ginneken B. Automatic lung segmentation from thoracic computed tomography scans using a hybrid approach with error detection. Med Phys. 2009;36:2934-2947.

    Not in our library10.1118/1.3147146

  24. 24.

    Mansoor A, Bagci U, Xu Z et al (2014) A generic approach to pathological lung segmentation. IEEE Trans Med Imaging 33:2293. 10.1109/TMI.2014.2337057

    Not in our library10.1109/tmi.2014.2337057

  25. 25.

    Zhang Y, Brady M, Smith S. Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Trans Med Imaging. 2001;20:45-57.

    Not in our library10.1109/42.906424

  26. 26.

    Rudyanto RD, Kerkstra S, van Rikxoort EM et al (2014) Comparing algorithms for automated vessel segmentation in computed tomography scans of the lung: the VESSEL12 study. Med Image Anal 18:1217–1232. 10.1016/J.MEDIA.2014.07.003

    Not in our library10.1016/j.media.2014.07.003

  27. 27.

    van Rikxoort EM, van Ginneken B, Kerkstra S (2011) LObe and Lung Analysis 2011 (LOLA11). https://lola11.grand-challenge.org.

    Not in our library

  28. 28.

    Hofmanninger J, Krenn M, Holzer M, Schlegl T, Prosch H, Langs G. Unsupervised identification of clinically relevant clusters in routine imaging data. 2016;192-200.

    Not in our library

  29. 29.

    Yushkevich PA, Piven J, Hazlett HC et al (2006) User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuroimage 31:1116–1128. 10.1016/J.NEUROIMAGE.2006.01.015

    Not in our library10.1016/j.neuroimage.2006.01.015

  30. 30.

    Ronneberger O, Fischer P, Brox T. U-net: Convolutional networks for biomedical image segmentation. 2015;234-241.

    Not in our library

  31. 31.

    Zhou X, Takayama R, Wang S, Hara T, Fujita H. Deep learning of the sectional appearances of 3D CT images for anatomical structure segmentation based on an FCN voting method. Med Phys. 2017;44:5221-5233.

    Not in our library10.1002/mp.12480

  32. 32.

    Isensee F, Petersen J, Kohl SAA, Jäger PF, Maier-Hein KH (2019) nnU-Net: breaking the spell on successful medical image segmentation. arXiv Prepr arXiv:1809.10486

    Not in our library

  33. 33.

    Ioffe S, Szegedy C. Batch normalization: accelerating deep network training by reducing internal covariate shift. 2015;448-456.

    Not in our library

  34. 34.

    Srivastava RK, Greff K, Schmidhuber J, Cortes C, Lawrence ND, Lee DD, et al. Training very deep networks. Advances in neural information processing systems. 2015;2377-2385.

    Not in our library

  35. 35.

    He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. 2016;770-778.

    Not in our library

  36. 36.

    Yu F, Koltun V (2015) Multi-scale context aggregation by dilated convolutions. arXiv Prepr arXiv:1511.07122

    Not in our library

  37. 37.

    Yu F, Koltun V, Funkhouser T (2017) Dilated residual networks. Proc IEEE Proceedings of the IEEE conference on computer vision and pattern recognition. pp 472–480. 10.1109/CVPR.2017.75

    Not in our library

  38. 38.

    Chen L-C, Papandreou G, Kokkinos I, Murphy K, Yuille AL. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Trans Pattern Anal Mach Intell. 2018;40:834-848.

    Not in our library10.1109/tpami.2017.2699184

  39. 39.

    Chest Imaging Platform (CIP). https://chestimagingplatform.org. Accessed Jun 8, 2020

    Not in our library

  40. 40.

    DeepMind (2018) Library to compute surface distance based performance metrics for segmentation tasks. https://github.com/deepmind/surface-distance.

    Not in our library

  41. 41.

    Guo F, Ng M, Goubran M, et al. Improving cardiac MRI convolutional neural network segmentation on small training datasets and dataset shift: a continuous kernel cut approach. Med Image Anal. 2020;61:101636.

    Not in our library10.1016/j.media.2020.101636

Cited by papers we hold 3

Reads

Read by people on this page

None yet.

Checked or read by AI through naulon

None yet.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.

Are you an author?

Is this your paper?

Claim it to add your note and be paid when agents cite it. Free.