Advertisement

3D Pulmonary Artery Segmentation from CTA Scans Using Deep Learning with Realistic Data Augmentation

  • Karen López-Linares RománEmail author
  • Isaac de La Bruere
  • Jorge Onieva
  • Lasse Andresen
  • Jakob Qvortrup Holsting
  • Farbod N. Rahaghi
  • Iván Macía
  • Miguel A. González Ballester
  • Raúl San José Estepar
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11040)

Abstract

The characterization of the vasculature in the mediastinum, more specifically the pulmonary artery, is of vital importance for the evaluation of several pulmonary vascular diseases. Thus, the goal of this study is to automatically segment the pulmonary artery (PA) from computed tomography angiography images, which opens up the opportunity for more complex analysis of the evolution of the PA geometry in health and disease and can be used in complex fluid mechanics models or individualized medicine. For that purpose, a new 3D convolutional neural network architecture is proposed, which is trained on images coming from different patient cohorts. The network makes use a strong data augmentation paradigm based on realistic deformations generated by applying principal component analysis to the deformation fields obtained from the affine registration of several datasets. The network is validated on 91 datasets by comparing the automatic segmentations with semi-automatically delineated ground truths in terms of mean Dice and Jaccard coefficients and mean distance between surfaces, which yields values of 0.89, 0.80 and 1.25 mm, respectively. Finally, a comparison against a Unet architecture is also included.

Keywords

Pulmonary artery Deep learning CTA Convolutional neural network Segmentation 

References

  1. 1.
  2. 2.
    Ebrahimdoost, Y., Qanadli, S.D., Nikravanshalmani, A., Ellis, T.J., Shojaee, Z.F., Dehmeshki, J.: Automatic segmentation of pulmonary artery (PA) in 3D pulmonary CTA images. In: Proceedings of the DSP, pp. 1–5 (2011)Google Scholar
  3. 3.
    Ibragimov, B., Toesca, D., Chang, D., Koong, A., Xing, L.: Combining deep learning with anatomical analysis for segmentation of the portal vein for liver SBRT planning. Phys. Med. Biol. 62(23), 8943–8958 (2017)CrossRefGoogle Scholar
  4. 4.
    Collins, J., Stern, E.J.: Chest Radiology, the Essentials. Lippincott Williams & Wilkins, Philadelphia (2007)Google Scholar
  5. 5.
    Jégou, S., Drozdzal, M., Vázquez, D., Romero, A., Bengio, Y.: The one hundred layers tiramisu: fully convolutional densenets for semantic segmentation. CoRR abs/1611.09326 (2016)Google Scholar
  6. 6.
    Linguraru, M.G., Pura, J.A., Uitert, R.L.V., Mukherjee, N., Summers, R.M.: Segmentation and quantification of pulmonary artery for noninvasive CT assessment of sickle cell secondary pulmonary hypertension. Med. Phys. 37(4), 1522–1532 (2010)CrossRefGoogle Scholar
  7. 7.
    Meijs, M., Manniesing, R.: Artery and vein segmentation of the cerebral vasculature in 4D CT using a 3D fully convolutional neural network. In: Proceedings of the SPIE, vol. 10575, p. 6 (2018)Google Scholar
  8. 8.
    Milletari, F., Navab, N., Ahmadi, S.: V-Net: fully convolutional neural networks for volumetric medical image segmentation. In: Proceedings of the 3DV, pp. 565–571 (2016)Google Scholar
  9. 9.
    Moses, D., Sammut, C., Zrimec, T.: Automatic segmentation and analysis of the main pulmonary artery on standard post-contrast CT studies using iterative erosion and dilation. Int. J. Comput. Assist. Radiol. Surg. 11(3), 381–395 (2016)CrossRefGoogle Scholar
  10. 10.
    Nardelli, P., Jimenez-Carretero, D., Bermejo-Peláez, D., Ledesma-Carbayo, M.J., Rahaghi, F.N., Estépar, R.S.J.: Deep-learning strategy for pulmonary artery-vein classification of non-contrast ct images. In: Proceedings of the ISBI, pp. 384–387 (2017)Google Scholar
  11. 11.
    Paszke, A., Chaurasia, A., Kim, S., Culurciello, E.: ENet: a deep neural network architecture for real-time semantic segmentation. CoRR abs/1606.02147 (2016)Google Scholar
  12. 12.
    Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation. CoRR abs/1505.04597 (2015)Google Scholar
  13. 13.
    Schroeder, W., Martin, K., Lorensen, B.: The Visualization Toolkit–An Object-Oriented Approach to 3D Graphics, 4th edn. Kitware Inc., Clifton Park (2006)Google Scholar
  14. 14.
    Sebbe, R., Gosselin, B., Coche, E.: Segmentation of opacified thorax vessels using model-driven active contour. In: Proceedings of the IEEE EMBS, vol. 3, pp. 2535–2538 (2005)Google Scholar
  15. 15.
    Truong, Q.A., et al.: Reference values for normal pulmonary artery dimensions by noncontrast cardiac computed tomography: the Framingham heart study. Circ. Cardiovasc. Imaging 5(1), 147–154 (2012)CrossRefGoogle Scholar
  16. 16.
    Yushkevich, P.A., et al.: User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuroimage 31(3), 1116–1128 (2006)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  • Karen López-Linares Román
    • 1
    • 2
    • 4
    Email author
  • Isaac de La Bruere
    • 3
  • Jorge Onieva
    • 4
  • Lasse Andresen
    • 4
  • Jakob Qvortrup Holsting
    • 4
  • Farbod N. Rahaghi
    • 3
  • Iván Macía
    • 1
  • Miguel A. González Ballester
    • 2
    • 5
  • Raúl San José Estepar
    • 4
  1. 1.Vicomtech Foundation and BiodonostiaSan SebastiánSpain
  2. 2.BCN Medtech, Universitat Pompeu FabraBarcelonaSpain
  3. 3.Division of Pulmonary and Critical Care MedicineBrigham and Women’s Hospital, Harvard Medical SchoolBostonUSA
  4. 4.Applied Chest Imaging Laboratory, Department of RadiologyBrigham and Women’s Hospital, Harvard Medical SchoolBostonUSA
  5. 5.ICREABarcelonaSpain

Personalised recommendations