Total Variation Random Forest: Fully Automatic MRI Segmentation in Congenital Heart Diseases

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10129)

Abstract

This paper proposes a fully automatic supervised segmentation technique for segmenting the great vessel and blood pool of pediatric cardiac MRIs of children with Congenital Heart Defects (CHD). CHD affects the overall anatomy of heart, rendering model-based segmentation framework infeasible, unless a large dataset of annotated images is available. However, the cardiac anatomy still retains distinct appearance patterns, which has been exploited in this work. In particular, Total Variation (TV) is introduced for solving the 3D disparity and noise removal problem. This results in homogeneous appearances within anatomical structures which is exploited further in a Random Forest framework. Context-aware appearance models are learnt using Random Forest (RF) for appearance-based prediction of great vessel and blood pool of an unseen subject during testing. We have obtained promising results on the HVSMR16 training dataset in a leave-one-out cross-validation.

Keywords

Total Variation Random Forest Congenital heart disease 3D cardiac MRI Automatic segmentation 

References

  1. 1.
    Assen, H.C., Danilouchkine, M.G., Behloul, F., Lamb, H.J., Geest, R.J., Reiber, J.H.C., Lelieveldt, B.P.F.: Cardiac LV segmentation using a 3D active shape model driven by fuzzy inference. In: Ellis, R.E., Peters, T.M. (eds.) MICCAI 2003. LNCS, vol. 2878, pp. 533–540. Springer, Heidelberg (2003). doi:10.1007/978-3-540-39899-8_66 CrossRefGoogle Scholar
  2. 2.
    Chan, S.H., et al.: An augmented Lagrangian method for total variation video restoration. IEEE Trans. Image Process. 20(11), 3097–3111 (2011)MathSciNetCrossRefGoogle Scholar
  3. 3.
    Delong, A., et al.: Fast approximate energy minimization with label costs. Int. J. Comput. Vision 96(1), 1–27 (2012)MathSciNetCrossRefMATHGoogle Scholar
  4. 4.
  5. 5.
  6. 6.
    Mukhopadhyay, A., Oksuz, I., Bevilacqua, M., Dharmakumar, R., Tsaftaris, S.A.: Unsupervised myocardial segmentation for cardiac MRI. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 12–20. Springer, Heidelberg (2015). doi:10.1007/978-3-319-24574-4_2 CrossRefGoogle Scholar
  7. 7.
    Pace, D.F., Dalca, A.V., Geva, T., Powell, A.J., Moghari, M.H., Golland, P.: Interactive whole-heart segmentation in congenital heart disease. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 80–88. Springer, Heidelberg (2015). doi:10.1007/978-3-319-24574-4_10 CrossRefGoogle Scholar
  8. 8.
    Vercauteren, T., Pennec, X., Perchant, A., Ayache, N.: Symmetric log-domain diffeomorphic registration: a demons-based approach. In: Metaxas, D., Axel, L., Fichtinger, G., Székely, G. (eds.) MICCAI 2008. LNCS, vol. 5241, pp. 754–761. Springer, Heidelberg (2008). doi:10.1007/978-3-540-85988-8_90 CrossRefGoogle Scholar
  9. 9.
    Zhuang, X.: Challenges and methodologies of fully automatic whole heart segmentation: a review. J. Healthc. Eng. 4(3), 371–408 (2013)CrossRefGoogle Scholar
  10. 10.
    Zikic, D., et al.: Decision forests for tissue-specific segmentation of high-grade gliomas in multi-channel MR. In: Ayache, N., Delingette, H., Golland, P., Mori, K. (eds.) MICCAI 2012. LNCS, vol. 7512, pp. 369–376. Springer, Heidelberg (2012). doi:10.1007/978-3-642-33454-2_46 CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  1. 1.Zuse Institute BerlinBerlinGermany

Personalised recommendations