Combination of Polyaffine Transformations and Supervised Learning for the Automatic Diagnosis of LV Infarct

  • Marc-Michel RohéEmail author
  • Nicolas Duchateau
  • Maxime Sermesant
  • Xavier Pennec
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9534)


In this article, we present an application of the polyaffine transformations to classify a population of hearts with myocardial infarction. Polyaffine transformations aim at representing motion by the combination of a limited number of affine transformations defined locally on a regional division of the space. We show that these transformations not only serve as a first (non-learnt) dimension reduction, but also allow to interpret each of the parameters and relate them to known clinical parameters. Then, we use standard supervised learning algorithms on these parameters to classify the population between infarcted and non-infarcted subjects. The method is applied on the STACOM statistical shape modeling labeled data consisting of 200 cases, comprising the same number of healthy subjects and patients with infarct. We train classifiers using different standard machine learning algorithms. Finally, we validate our method with 10-fold cross-validation and get more than 95 % of correct classification on yet-unseen data. The method is promising and ready to be tested on the remaining 200 test cases of the challenge.



The authors acknowledge the partial funding by the EU FP7-funded project MD-Paedigree (Grant Agreement 600932).


  1. 1.
    Konstam, M.A., Kramer, D.G., Patel, A.R., Maron, M.S., Udelson, J.E.: Left ventricular remodeling in heart failure: current concepts in clinical significance and assessment. JACC: Cardiovasc. Imaging 4, 98–108 (2011)Google Scholar
  2. 2.
    Bijnens, B., Claus, P., Weidemann, F., Strotmann, J., Sutherland, G.: Investigating cardiac function using motion and deformation analysis in the setting of coronary artery disease. Circulation 116, 2453–2464 (2007)CrossRefGoogle Scholar
  3. 3.
    Sudarshan, V., Acharya, U.R., Yin-Kwee Ng, E., Meng, C.S., Tan, R.S., Ghista, D.N.: Automated identification of infarcted myocardium tissue characterisation using ultrasound images: a review. IEEE Rev. Biomed. Eng. 8, 86–97 (2013)CrossRefGoogle Scholar
  4. 4.
    Fonseca, C., Backhaus, M., Bluemke, D., Britten, R., Chung, J., Cowan, B., Dinov, I., Finn, J., Hunter, P., Kadish, A., Lee, D., Lima, J., Medrano-Gracia, P., Shivkumar, K., Suinesiaputra, A., Tao, W., Young, A.: The cardiac atlas project-an imaging database for computational modeling and statistical atlases of the heart. Bioinformatics 27(5), 2288–2295 (2011). 2011–08-15 00:00:00.0CrossRefGoogle Scholar
  5. 5.
    Zhang, X., et al.: Orthogonal shape modes describing clinical indices of remodeling. In: van Assen, H., Bovendeerd, P., Delhaas, T. (eds.) FIMH 2015. LNCS, vol. 9126, pp. 273–281. Springer, Heidelberg (2015)CrossRefGoogle Scholar
  6. 6.
    Arsigny, V., Commowick, O., Ayache, N., Pennec, X.: A fast and log-euclidean polyaffine framework for locally linear registration. J. Math. Imaging Vis. 33, 222–238 (2009)MathSciNetCrossRefGoogle Scholar
  7. 7.
    McLeod, K., Sermesant, M., Beerbaum, P., Pennec, X.: Spatio-temporal tensor decomposition of a polyaffine motion model for a better analysis of pathological left ventricular dynamics. IEEE Trans. Med. Imaging 34, 1562–1575 (2015)CrossRefGoogle Scholar
  8. 8.
    Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., Duchesnay, E.: Scikit-learn: Machine learning in Python. J. Mach. Learn. Res. 12, 2825–2830 (2011)MathSciNetzbMATHGoogle Scholar
  9. 9.
    Rosipal, R., Krämer, N.C.: Overview and recent advances in partial least squares. In: Saunders, C., Grobelnik, M., Gunn, S., Shawe-Taylor, J. (eds.) SLSFS 2005. LNCS, vol. 3940, pp. 34–51. Springer, Heidelberg (2006)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  • Marc-Michel Rohé
    • 1
    Email author
  • Nicolas Duchateau
    • 1
  • Maxime Sermesant
    • 1
  • Xavier Pennec
    • 1
  1. 1.Inria Sophia-Antipolis, Asclepios Research GroupSophia-AntipolisFrance

Personalised recommendations