Advertisement

MRI Whole Heart Segmentation Using Discrete Nonlinear Registration and Fast Non-local Fusion

  • Mattias P. Heinrich
  • Julien Oster
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10663)

Abstract

We present a robust and accurate method for multi-atlas segmentation of whole heart MRI scans. After preprocessing, which includes resampling to isotropic voxel sizes and cropping or padding to same dimensions, all training scans are registered linearly and nonlinearly to an unseen set of test scans. We employ the efficient discrete registration framework called deeds that captures large shape variations across scans, performed best in a recent registration comparison on abdominal scans and requires less than 2 min of computation time per scan. Subsequently, we perform multi-atlas label fusion using a non-local means approach with a normalised SSD metric and a fast implementation using boxfilters. Subsequently, a multi-label random walk is performed on the obtained probability maps for an edge-preserving smoothing. Without performing any domain-specific parameter tuning, we obtained a Dice accuracy of 86.0% (averaged across 7 labels) and 87.0% for the whole heart on the MRI test dataset, which is the first rank of the MICCAI 2017 challenge. The segmentations are also visually very smooth using this fully automatic method.

Notes

Acknowledgements

We would like to thank the organisers of the MM-WHS 2017 for providing this rich new dataset to the public, which enables the evaluation of new algorithms for the problem of detailed 3D heart segmentation.

References

  1. 1.
    Grothues, F., Smith, G.C., Moon, J.C., Bellenger, N.G., Collins, P., Klein, H.U., Pennell, D.J.: Comparison of interstudy reproducibility of cardiovascular magnetic resonance with 2D echocardiography in normal subjects and in patients with heart failure or left ventricular hypertrophy. Am. J. Cardiol. 90(1), 29–34 (2002)CrossRefGoogle Scholar
  2. 2.
    Ramanathan, C., Ghanem, R.N., Jia, P., Ryu, K., Rudy, Y.: Noninvasive electrocardiographic imaging for cardiac electrophysiology and arrhythmia. Nat. Med. 10(4), 422–428 (2004)CrossRefGoogle Scholar
  3. 3.
    Vuissoz, P.A., Odille, F., Fernandez, B., Lohezic, M., Benhadid, A., Mandry, D., Felblinger, J.: Free-breathing imaging of the heart using 2D cine-GRICS with assessment of ventricular volumes and function. J. Magn. Reson Imaging 35(2), 340–351 (2012)CrossRefGoogle Scholar
  4. 4.
    Nazarian, S., Bluemke, D.A., Lardo, A.C., Zviman, M.M., Watkins, S.P., Dickfeld, T.L., Meininger, G.R., Roguin, A., Calkins, H., Tomaselli, G.F., et al.: Magnetic resonance assessment of the substrate for inducible ventricular tachycardia in nonischemic cardiomyopathy. Circulation 112(18), 2821–2825 (2005)CrossRefGoogle Scholar
  5. 5.
    Nielles-Vallespin, S., Mekkaoui, C., Gatehouse, P., Reese, T.G., Keegan, J., Ferreira, P.F., Collins, S., Speier, P., Feiweier, T., Silva, R., et al.: In vivo diffusion tensor MRI of the human heart: reproducibility of breath-hold and navigator-based approaches. Magn. Reson. Med. 70(2), 454–465 (2013)CrossRefGoogle Scholar
  6. 6.
    Tobon-Gomez, C., Geers, A.J., Peters, J., Weese, J., Pinto, K., Karim, R., Ammar, M., Daoudi, A., Margeta, J., Sandoval, Z., et al.: Benchmark for algorithms segmenting the left atrium from 3D CT and MRI datasets. IEEE Trans. Med. Imag. 34(7), 1460–1473 (2015)CrossRefGoogle Scholar
  7. 7.
    Kutra, D., Saalbach, A., Lehmann, H., Groth, A., Dries, S.P.M., Krueger, M.W., Dössel, O., Weese, J.: Automatic multi-model-based segmentation of the left atrium in cardiac MRI scans. In: Ayache, N., Delingette, H., Golland, P., Mori, K. (eds.) MICCAI 2012. LNCS, vol. 7511, pp. 1–8. Springer, Heidelberg (2012).  https://doi.org/10.1007/978-3-642-33418-4_1 CrossRefGoogle Scholar
  8. 8.
    Zhuang, X., Rhode, K.S., Razavi, R.S., Hawkes, D.J., Ourselin, S.: A registration-based propagation framework for automatic whole heart segmentation of cardiac MRI. IEEE Trans. Med. Imag. 29(9), 1612–1625 (2010)CrossRefGoogle Scholar
  9. 9.
    Zhuang, X., Shen, J.: Multi-scale patch and multi-modality atlases for whole heart segmentation of MRI. Med. Imag. Anal. 31, 77–87 (2016)CrossRefGoogle Scholar
  10. 10.
    Wolterink, J.M., Leiner, T., Viergever, M.A., Išgum, I.: Dilated convolutional neural networks for cardiovascular MR segmentation in congenital heart disease. In: Zuluaga, M.A., Bhatia, K., Kainz, B., Moghari, M.H., Pace, D.F. (eds.) RAMBO/HVSMR -2016. LNCS, vol. 10129, pp. 95–102. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-52280-7_9 CrossRefGoogle Scholar
  11. 11.
    Oktay, O., Ferrante, E., Kamnitsas, K., Heinrich, M., Bai, W., Caballero, J., Guerrero, R., Cook, S., de Marvao, A., O’Regan, D., et al.: Anatomically constrained neural networks (ACNN): Application to cardiac image enhancement and segmentation. arXiv preprint arXiv:1705.08302 (2017)
  12. 12.
    Heinrich, M., Jenkinson, M., Brady, J., Schnabel, J.: MRF-based deformable registration and ventilation estimation of lung CT. IEEE Trans. Med. Imag. 32(7), 1239–1248 (2013)CrossRefGoogle Scholar
  13. 13.
    Xu, Z., Lee, C., Heinrich, M., Modat, M., Rueckert, D., Ourselin, S., Abramson, R., Landman, B.: Evaluation of six registration methods for the human abdomen on clinically acquired CT. IEEE Trans. Biomed. Eng. 1–10 (2016)Google Scholar
  14. 14.
    Heinrich, M.P., Jenkinson, M., Papież, B.W., Brady, S.M., Schnabel, J.A.: Towards realtime multimodal fusion for image-guided interventions using self-similarities. In: Mori, K., Sakuma, I., Sato, Y., Barillot, C., Navab, N. (eds.) MICCAI 2013. LNCS, vol. 8149, pp. 187–194. Springer, Heidelberg (2013).  https://doi.org/10.1007/978-3-642-40811-3_24 CrossRefGoogle Scholar
  15. 15.
    Wang, H., Suh, J.W., Das, S.R., Pluta, J.B., Craige, C., Yushkevich, P.A.: Multi-atlas segmentation with joint label fusion. IEEE Trans. Patt. Anal. Mach. Intell. 35(3), 611–623 (2013)CrossRefGoogle Scholar
  16. 16.
    Asman, A.J., Landman, B.A.: Non-local statistical label fusion for multi-atlas segmentation. Med. Imag. Anal. 17(2), 194–208 (2013)CrossRefGoogle Scholar
  17. 17.
    Heinrich, M.P., Simpson, I., Papież, B., Brady, J., Schnabel, J.: Deformable image registration by combining uncertainty estimates from supervoxel belief propagation. Med. Imag. Anal. 27, 57–71 (2016)CrossRefGoogle Scholar
  18. 18.
    Coupé, P., Manjón, J.V., Fonov, V., Pruessner, J., Robles, M., Collins, D.L.: Patch-based segmentation using expert priors: application to hippocampus and ventricle segmentation. NeuroImage 54(2), 940–954 (2011)CrossRefGoogle Scholar
  19. 19.
    Heinrich, M.P., Papież, B.W., Schnabel, J.A., Handels, H.: Non-parametric discrete registration with convex optimisation. In: Ourselin, S., Modat, M. (eds.) WBIR 2014. LNCS, vol. 8545, pp. 51–61. Springer, Cham (2014).  https://doi.org/10.1007/978-3-319-08554-8_6 Google Scholar
  20. 20.
    Langerak, T., Van Der Heide, U., Kotte, A., Viergever, M., Van Vulpen, M., Pluim, J.: Label fusion in atlas-based segmentation using a selective and iterative method for performance level estimation (SIMPLE). IEEE Trans. Med. Imag. 29(12), 2000–2008 (2010)CrossRefGoogle Scholar
  21. 21.
    Xu, Z., Asman, A.J., Shanahan, P.L., Abramson, R.G., Landman, B.A.: SIMPLE is a good idea (and better with context learning). In: Golland, P., Hata, N., Barillot, C., Hornegger, J., Howe, R. (eds.) MICCAI 2014. LNCS, vol. 8673, pp. 364–371. Springer, Cham (2014).  https://doi.org/10.1007/978-3-319-10404-1_46 Google Scholar
  22. 22.
    Grady, L.: Multilabel random walker image segmentation using prior models. In: CVPR, pp. 763–770 (2005)Google Scholar
  23. 23.
    Heinrich, M.P., Blendowski, M.: Multi-organ segmentation using vantage point forests and binary context features. In: Ourselin, S., Joskowicz, L., Sabuncu, M.R., Unal, G., Wells, W. (eds.) MICCAI 2016. LNCS, vol. 9901, pp. 598–606. Springer, Cham (2016).  https://doi.org/10.1007/978-3-319-46723-8_69 CrossRefGoogle Scholar
  24. 24.
    Krähenbühl, P., Koltun, V.: Efficient inference in fully connected CRFs with Gaussian edge potentials. In: Proceedings of NIPS, pp. 2–9 (2011)Google Scholar
  25. 25.
    Oguz, I., Kashyap, S., Wang, H., Yushkevich, P., Sonka, M.: Globally optimal label fusion with shape priors. In: Ourselin, S., Joskowicz, L., Sabuncu, M.R., Unal, G., Wells, W. (eds.) MICCAI 2016. LNCS, vol. 9901, pp. 538–546. Springer, Cham (2016).  https://doi.org/10.1007/978-3-319-46723-8_62 CrossRefGoogle Scholar
  26. 26.
    Bai, W., Shi, W., Ledig, C., Rueckert, D.: Multi-atlas segmentation with augmented features for cardiac MR images. Med. Imag. Anal. 19(1), 98–109 (2015)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Institute of Medical InformaticsUniversity of LübeckLübeckGermany
  2. 2.IADI, U947, Inserm, CHRU de NancyVandoeuvre les NancyFrance

Personalised recommendations