International Journal of Computer Vision

, Volume 46, Issue 2, pp 103–128 | Cite as

Cluster Analysis of Biomedical Image Time-Series

  • Axel Wismüller
  • Oliver Lange
  • Dominik R. Dersch
  • Gerda L. Leinsinger
  • Klaus Hahn
  • Benno Pütz
  • Dorothee Auer

Abstract

In this paper, we present neural network clustering by deterministic annealing as a powerful strategy for self-organized segmentation of biomedical image time-series data identifying groups of pixels sharing common properties of local signal dynamics. After introducing the theoretical concept of minimal free energy vector quantization and related clustering techniques, we discuss its potential to serve as a multi-purpose computer vision strategy to image time-series analysis and visualization for many fields of medicine ranging from biomedical basic research to clinical assessment of patient data. In particular, we present applications to (i) functional MRI data analysis for human brain mapping, (ii) dynamic contrast-enhanced perfusion MRI for the diagnosis of cerebrovascular disease, and (iii) magnetic resonance mammography for the analysis of suspicious lesions in patients with breast cancer. This wide scope of completely different medical applications illustrates the flexibility and conceptual power of neural network vector quantization in this context. Although there are obvious methodological similarities, each application requires specific careful consideration w.r.t. data preprocessing, postprocessing and interpretation. This challenge can only be managed by close interdisciplinary cooperation of medical doctors, engineers, and computer scientists. Hence, this field of research can serve as an example for lively cross-fertilization between computer vision and related research.

clustering time-series neural networks deterministic annealing medical imaging functional MRI dynamic perfusion MRI MRI mammography 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Anderberg, M.R. (Ed.). 1973. Cluster Analysis for Applications. Academic Press: New York.Google Scholar
  2. Axel, L. 1980. Cerebral blood flow determination by rapid-sequence computed tomography. Radiology, 137:679–686.Google Scholar
  3. Bandettini, P.A., Jesmanowicz, A., Wong, E.C., and Hyde, J.S. 1993. Processing strategies for time-course data sets in functional MRI of the human brain. Magn. Reson. Med., 30:161–173.Google Scholar
  4. Baumgartner, R., Windischberger, C., and Moser, E. 1998. Quantification in functional magnetic resonance imaging: Fuzzy clustering vs. correlation analysis. Magnetic Resonance Imaging, 16(2):115–125.Google Scholar
  5. Becker, S. 1996. Mutal information maximization: Models of cortical self-organization. Network, 7:7–31.Google Scholar
  6. Bell, A.J. and Sejnowski, T.J. 1995. An information-maximization approach to blind separation and blind deconvolution. Neural Computation, 7:1129–1159.Google Scholar
  7. Bezdek, J.C. 1981. Pattern Recognition with Fuzzy Objective Function Algorithms. Plenum Press: New York.Google Scholar
  8. Bishop, C.M., Svensen, M., and Williams, C.K.I. 1998a. Developments of the generative topographic mapping. Technical Report NCRG-98-012, Neural Computing Research Group, Aston University, Birmingham.Google Scholar
  9. Bishop, C.M., Svensen, M., and Williams, C.K.I. 1998b. GTM: The generative topographic mapping. Neural Computation, 10(1):215–234.Google Scholar
  10. Chuang, K.H., Chiu, M.J., Lin, C.C., and Chen, J.H. 1999. Model-free functional MRI analysis using Kohonen clustering neural network and fuzzy c-means. IEEE Transactions on Medical Imaging, 18(12):1117–1128.Google Scholar
  11. Dempster, A.P., Laird, N.M., and Rubin, D.B. 1977. Maximum likelihood from incomplete data via the EMalgorithm. J. Royal. Statist. Soc. Ser. B (methodological), 39:1–38.Google Scholar
  12. Dersch, D.R. 1996. Eigenschaften neuronaler Vektorquantisierer und ihre Anwendung in der Sprachverarbeitung. Verlag Harri Deutsch, Reihe Physik, Bd. 54, Thun, Frankfurt am Main. ISBN 3-8171-1492-3.Google Scholar
  13. Dersch, D.R., Albrecht, S., and Tavan, P. 1996. Hierarchical fuzzy clustering. In Symposion über biologische Informationsver-arbeitung und Neuronale Netze – SINN’ 95, Konferenzband, A. Wismüller and D.R. Dersch (Eds.). Hanns-Seidel-Stiftung, München.Google Scholar
  14. Dersch, D.R. and Tavan, P. 1994a. Control of annealing in minimal free energy vector quantization. In Proceedings of the IEEE International Conference on Neural Networks ICNN'94, Orlando, Florida, pp. 698–703.Google Scholar
  15. Dersch, D.R. and Tavan, P. 1994b. Load balanced vector quantization. In Proceedings of the International Conference on Artificial Neural Networks ICANN, Springer, pp. 1067–1070.Google Scholar
  16. Dersch, D.R. and Tavan, P. 1995. Asymptotic level density in topological feature maps. IEEE Transactions on Neural Networks, 6(1):230–236.Google Scholar
  17. Ding, X., Masaryk, T., Ruggieri, P., and Tkach, J. 1996. Detection of activation patterns in dynamic functional MRI with a clustering technique. In Proc., SMR, 4th Annual Meeting, New York, p. 1798.Google Scholar
  18. Ding, X., Tkach, J., Ruggieri, P., and Masaryk, T. 1994. Analysis of time-course functional mri data with clustering method without use of reference signal. In Proc., SMR, 2nd Annual Meeting, San Francisco, p. 630.Google Scholar
  19. Duda, R.O. and Hart, P.E. 1973. Pattern Classification and Scene Analysis. Wiley: New York.Google Scholar
  20. Erwin, E., Obermayer, K., and Schulten, K. 1992a. Self-organizing maps: Ordering, convergence properties, and energy functions. Biological Cybernetics, 61:47–55.Google Scholar
  21. Erwin, E., Obermayer, K., and Schulten, K. 1992b. Self-organizing maps: Stationary states, metastability, and convergence rate. Biological Cybernetics, 61:35–45.Google Scholar
  22. Fischer, H. and Hennig, J. 1999. Neural-network based analysis of MR time series. Magn. Reson. Med., 41(1):124–131.Google Scholar
  23. Forgy, E.W. 1965. Cluster analysis of multivariate data: Efficiency vs. interpretability of classifications. Biometrics, 21:768.Google Scholar
  24. Forman, S.D., Cohen, J.D., Fitzgerald, M., Eddy, W.F., Mintun, M.A., and Noll, D.C. 1995. Improved assessment of significant activation in functional magnetic resonance imaging (fMRI): Use of a cluster-size threshold. Magn. Reson. Med., 33:636–647.Google Scholar
  25. Galicki, M., Möller, U., and Witte, H. 1997. Neural clustering networks based on global optimisation of prototypes in metric spaces. Neural Computing and Applications, 5:2–13.Google Scholar
  26. Golay, X., Kollias, S., Meier, D., and Boesinger, P. 1996. Optimization of a fuzzy clustering technique and comparison with conventional post processing methods in fMRI. In Proc., SMR, 4th Annual Meeting, New York, p. 1787.Google Scholar
  27. Goutte, C., Toft, P., Rostrup, E., Nielsen, F.A., and Hansen, L.K. 1999. On clustering fMRI time series. Neuroimage, 9:298–310.Google Scholar
  28. Graepel, T., Burger, M., and Obermayer, K. 1997. Phase transitions in stachastic self-organizing maps. Physical Review E, 56(4):3876–3890.Google Scholar
  29. Graepel, T., Burger, M., and Obermayer, K. 1998. Self-organizing maps: Generalizations and new optimization techniques. Neuro-computing, 20:173–190.Google Scholar
  30. Hyvärinen, A. 1999. Survey on independent component analysis. Neural Computing Surveys, 2:94–128.Google Scholar
  31. Kloppenburg, M. 1996. Lokale Hauptkomponentenanalyse in hoch-dimensionalen Merkmalsräumen—Methoden zur Datenreduktion. Diplomarbeit, Ludwig-Maximilians-Universität, München.Google Scholar
  32. Kloppenburg, M. and Tavan, P. 1997. Deterministic annealing for density estimation by multivariate normal mixtures. Phys. Rev. E, 55:2089–2092.Google Scholar
  33. Kohonen, T. 1989. Self-Organisation and Associative Memory. Springer: Berlin.Google Scholar
  34. Kohonen, T. 1990. The self-organizing map. Proceedings of the IEEE, 78(9):1464–1480.Google Scholar
  35. Kohonen, T., Hynninen, J., Kangas, J., and Laaksonen, J. 1996. SOM PAK: The self-organizing map program package. Helsinki University of Technology, Laboratory of Computer and Information Science, Rakentajanaukio 2 C, SF-02150 Espoo, Finland.Google Scholar
  36. Linde, Y., Buzo, A., and Gray, R.M. 1980. An algorithm for vector quantizer design. IEEE Transactions on Communications, 28:84–95.Google Scholar
  37. Martinetz, T.M. and Schulten, K. 1991. A'neural gas’ network learns topologies. In Proceedings of the International Conference on Artificial Neural Networks ICANN, Elsevier Science Publishers: Amsterdam, pp. 397–402.Google Scholar
  38. Martinetz, T.M. and Schulten, K. 1994. Topology representing networks. Neural Networks, 7:507–522.Google Scholar
  39. Milligan, G.W. and Cooper, M.C. 1985. An examination of procedures for determining the number of clusters in a data set. Psychometrika, 50:159–179.Google Scholar
  40. Moser, E., Teichtmeister, C., and Diemling, M. 1996. Reproducibility and postprocessing of gradient echo functional MRI to improve localization of brain activity in the human visual cortex. Magnetic Resonance Imaging, 14:567–579.Google Scholar
  41. Ogawa, S., Lee, T., Kay, A.R., and Tank, D.W. 1990. Brain magnetic-resonance-imaging with contrast dependent on blood oxygenation. Proc. Natl. Acad. Sci. USA, 87(24):9868–9872.Google Scholar
  42. Østergaard, L., Weisskopf, R.M., Chesler, D.A., Gyldensted, C., and Rosen, B.R. 1996. High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. Part I: Mathematical approach and statistical analysis. Magn. Res. Med., 36:715–725.Google Scholar
  43. Ritter, H. 1991. Asymptotic level density for a class of vector quantization processes. IEEE Transactions on Neural Networks, 1(2):173–175.Google Scholar
  44. Ritter, H., Martinetz, T., and Schulten, K. 1991a. Neural Networks. Addison Wesley: New York.Google Scholar
  45. Ritter, H., Martinetz, T., and Schulten, K. 1991b. Neuronale Netze. Addison Wesley: New York.Google Scholar
  46. Rose, K., Gurewitz, E., and Fox, G. 1990. A deterministic annealing approach to clustering. Pattern Recognition Letters, 11(11):589–594.Google Scholar
  47. Rose, K., Gurewitz, E., and Fox, G.C. 1992. Vector quantization by deterministic annealing. IEEE Transactions on Information Theory, 38(4):1249–1257.Google Scholar
  48. Rosen, B.R., Belliveau, J.W., Vevea, J.M., and Brady, T.J. 1990. Perfusion imaging with NMR contrast agents. Magn. Reson. Med., 14:249–265.Google Scholar
  49. Sammon, J.W. 1969. Anonlinear mapping for data structure analysis. IEEE Transactions on Computers, C 18:401–409.Google Scholar
  50. Scarth, G., McIntyre, M., Wowk, B., and Somorjai, R.L. 1995. Detection of novelty in functional images using fuzzy clustering. In Proc., SMR, 3rd Annual Meeting, Nice, p. 238.Google Scholar
  51. Scarth, G., Moser, E., Baumgartner, R., Alexander M., and Somorjai, R.L. 1996. Paradigm-free fuzzy clustering-detected activations in fMRI: A case study. In Proc., SMR, 4th Annual Meeting, New York, p. 1784.Google Scholar
  52. Weisskoff, R. 1998. Personal communication.Google Scholar
  53. Willshaw, D.J. and vonder Malsburg, C. 1976. How patterned neural connections can be set up by self-organization. Proceedings of the Royal Society, London, B 194:431–445.Google Scholar
  54. Wismüller, A. and Dersch, D.R. 1997. Neural network computation in biomedical research: Chances for conceptual cross-fertilization. Theory in Biosciences, 116(3):229–240.Google Scholar
  55. Wismüller, A., Dersch, D.R., Lipinski, B., Hahn, K., and Auer, D. 1998a. Hierarchical clustering of fMRI time-series by deterministic annealing. Neuroimage, 7(4):S593.Google Scholar
  56. Wismüller, A., Dersch, D.R., Lipinski, B., Hahn, K., and Auer, D. 1998b. Hierarchical unsupervised clustering of fMRI data by deterministic annealing. In Proceedings of the Sixth Scientific Meeting of the International Society of Magnetic Resnonace in Medicine 1998, Sydney, p. 412.Google Scholar
  57. Wismüller, A., Dersch, D.R., Lipinski, B., Hahn, K., and Auer, D. 1998c. A neural network approach to functional MRI pattern analysis—clustering of time-series by hierarchical vector quantization. In ICANN'98—Proceedings of the 8th International Conference on Artificial Neural Networks, Skövde, Sweden. Perspectives in Neural Computing, vol. 2, L. Niklasson, M. Bodèn, and T. Ziemke (Eds.). Springer-Verlag: London, Berlin, New York, pp. 123–128.Google Scholar
  58. Wismüller, A., Vietze, F., and Dersch, D.R. 2000. Segmentation with neural networks. In Handbook of Medical Imaging, I. Bankman, R. Rangayyan, A. Evans, R. Woods, E. Fishman, and H. Huang (Eds.). Johns Hopkins University, Baltimore, Academic Press: NewYork. ISBN 0120777908.Google Scholar
  59. Wismüller, A., Vietze, F., Dersch, D.R., Hahn, K., and Ritter, H. 1998. The deformable feature map—adaptive plasticity in function approximation. In ICANN'98—Proceedings of the 8th International Conference on Artificial Neural Networks, Skövde, Sweden. Perspectives in Neural Computing, vol. 1, L. Niklasson, M. Bodèn, and T. Ziemke (Eds.). Springer-Verlag: London, Berlin, New York, pp. 222–227.Google Scholar
  60. Woods, R.P., Cherry, S.R., and Mazziotta, J.C. 1992. Rapid automated algorithm for aligning and reslicing PET images. Journal of Computer Assisted Tomography, 16:620–633.Google Scholar
  61. Zierler, K.L. 1965. Theoretical basis of indicator-dilution methods for measuring flow and volume. Circ. Res., 16:393–407.Google Scholar

Copyright information

© Kluwer Academic Publishers 2002

Authors and Affiliations

  • Axel Wismüller
    • 1
  • Oliver Lange
    • 1
  • Dominik R. Dersch
    • 2
  • Gerda L. Leinsinger
    • 1
  • Klaus Hahn
    • 1
  • Benno Pütz
    • 3
  • Dorothee Auer
    • 3
  1. 1.Institut für Radiologische DiagnostikLudwig-Maximilians-Universität München, Klinikum InnenstadtMünchenGermany
  2. 2.Crux Cybernetics Corp.SydneyAustralia
  3. 3.Max Planck Institute of PsychiatryMunichGermany

Personalised recommendations