Analytical and Bioanalytical Chemistry

, Volume 387, Issue 5, pp 1801–1807 | Cite as

Multivariate feature selection and hierarchical classification for infrared spectroscopy: serum-based detection of bovine spongiform encephalopathy

  • Bjoern H. Menze
  • Wolfgang Petrich
  • Fred A. Hamprecht
Original Paper


A hierarchical scheme has been developed for detection of bovine spongiform encephalopathy (BSE) in serum on the basis of its infrared spectral signature. In the first stage, binary subsets between samples originating from diseased and non-diseased cattle are defined along known covariates within the data set. Random forests are then used to select spectral channels on each subset, on the basis of a multivariate measure of variable importance, the Gini importance. The selected features are then used to establish binary discriminations within each subset by means of ridge regression. In the second stage of the hierarchical procedure the predictions from all linear classifiers are used as input to another random forest that provides the final classification. When applied to an independent, blinded validation set of 160 further spectra (84 BSE-positives, 76 BSE-negatives), the hierarchical classifier achieves a sensitivity of 92% and a specificity of 95%. Compared with results from an earlier study based on the same data, the hierarchical scheme performs better than linear discriminant analysis with features selected by genetic optimization and robust linear discriminant analysis, and performs as well as a neural network and a support vector machine.


Diagnostic pattern recognition Random forest Gini importance Feature selection Hierarchical classification 



The authors acknowledge the contributions of W. Köhler, T. Martin, and J. Möcks, and partial financial support under grant no. HA-4364 from the DFG (German National Science Foundation) and the Robert Bosch GmbH.


  1. 1.
    Gremlich H-U, Yan B (eds) (2001) Infrared and Raman spectroscopy of biological materials, vol 24 of Practical spectroscopy series. Marcel Dekker, New YorkGoogle Scholar
  2. 2.
    Morris MD, Berger A, Mahadevan-Jansen A (eds) (2005) J Biomed Opt 10:031101–031119Google Scholar
  3. 3.
    Naumann D (2001) Appl Spectrosc Rev 36:198–238CrossRefGoogle Scholar
  4. 4.
    Petrich W (2001) Appl Spectrosc Rev 36:181–237CrossRefGoogle Scholar
  5. 5.
    Chalmers JM, Griffiths PR (eds) (2002) Handbook of vibrational spectroscopy, vol 5. Wiley, ChichesterGoogle Scholar
  6. 6.
    Beleites C, Steiner G, Sowa MG, Baumgartner R, Sobottka S, Schackert G, Salzer R (2005) Vib Spectrosc 38:143–149CrossRefGoogle Scholar
  7. 7.
    Lasch P, Schmitt J, Beekes M, Udelhoven T, Eiden M, Fabian H, Petrich W, Naumann D (2003) Anal Chem 75:6673–6678CrossRefGoogle Scholar
  8. 8.
    Petrich W, Dolenko B, Früh J, Ganz M, Greger H, Jacob S, Keller F, Nikulin AE, Otto M, Quarder O, Somorjai RL, Staib A, Werner G, Wielinger H (2000) Appl Optics 39:3372–3379CrossRefGoogle Scholar
  9. 9.
    Staib A, Dolenko B, Fink DJ, Früh J, Nikulin AE, Otto M, Pessin-Minsley MS, Quarder O, Somorjai RL, Thienel U, Werner G, Petrich W (2001) Clin Chim Acta 308:79–89CrossRefGoogle Scholar
  10. 10.
    Himmelreich U, Somorjai RL, Dolenko B, Lee OC, Daniel HM, Murray R, Mountford CE, Sorrell TC (2003) Appl Environ Microbiol 69:4566–4574CrossRefGoogle Scholar
  11. 11.
    Breiman L (2001) Mach Learn J 45:5–32CrossRefGoogle Scholar
  12. 12.
    Martin TC, Moecks J, Belooussov A, Cawthraw S, Dolenko B, Eiden M, Von Frese J, Kohler W, Schmitt J, Somorjai RL, Udelhoven T, Verzakov S, Petrich W (2004) Analyst 129:897–901CrossRefGoogle Scholar
  13. 13.
    Myles A, Feudale R, Liu Y, Woody N, Brown S (2004) J Chemom 18:275–285CrossRefGoogle Scholar
  14. 14.
    Svetnik V, Liaw A, Tong C, Culberson JC, Sheridan RP, Feuston BP (2003) J Chem Inf Comput Sci 43:1947–1958CrossRefGoogle Scholar
  15. 15.
    Meyer D, Leisch F, Hornik K (2003) Neurocomputing 55:169–186CrossRefGoogle Scholar
  16. 16.
    Li S, Fedorowicz A, Singh H, Soderholm SC (2005) J Chem Inf Model 45:952–964CrossRefGoogle Scholar
  17. 17.
    Diaz-Uriarte R, Alvarez de Andres S (2006) BMC Bioinform 7Google Scholar
  18. 18.
    Jiang H, Deng Y, Chen H-S, Tao L, Sha Q, Chen J, Tsai C-J, Zhang S (2004) BMC Bioinform 5Google Scholar
  19. 19.
    Hastie T, Tibshirani R, Friedman J (2001) The elements of statistical learning. Springer series in statistics. Springer, Berlin Heidelberg New YorkGoogle Scholar
  20. 20.
    Wolpert DH (1992) Neural Netw 5:241–259CrossRefGoogle Scholar
  21. 21.
    Schmitt J, Udelhoven T (2001) Use of artificial neural networks in biomedical diagnostics. In: Gremlich H-U, Yan B (eds) Infrared and Raman spectroscopy of biological materials, vol 24 of practical spectroscopy series. Marcel Dekker, New York, pp 379–420Google Scholar
  22. 22.
    Maquelin K, Kirschner C, Choo-Smith LP, Ngo-Thi NA, Van Vreeswijk T, Stämmler M, Endtz HP, Bruining D, Naumann HA, Puppels GJ (2003) J Clin Microbiol 41:324–329CrossRefGoogle Scholar
  23. 23.
    Liaw A, Wiener M (2002) R News 2:18–22Google Scholar
  24. 24.
    Ihaka R, Gentleman R (1996) J Comput Graph Stat 5:299–314CrossRefGoogle Scholar
  25. 25.
    Schmitt J, Lasch P, Beekes M, Udelhoven T, Eiden M, Fabian H, Petrich W, Naumann D (2004) Proc SPIE 5321:36–43CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2007

Authors and Affiliations

  • Bjoern H. Menze
    • 1
    • 2
  • Wolfgang Petrich
    • 2
    • 3
  • Fred A. Hamprecht
    • 1
    • 2
  1. 1.Interdisciplinary Center for Scientific Computing (IWR)University of HeidelbergHeidelbergGermany
  2. 2.Department of Physics and AstronomyUniversity of HeidelbergHeidelbergGermany
  3. 3.Roche Diagnostics GmbHMannheimGermany

Personalised recommendations