Skip to main content
SpringerLink
Log in

Combining rules using local statistics and uncertainty estimates for improved ensemble segmentation

  • Representation, Processing, Analysis, and Understanding of Images
  • Published:
Pattern Recognition and Image Analysis Aims and scope Submit manuscript

An Erratum to this article was published on 01 January 2018

This article has been updated

Abstract

Segmentation using an ensemble of classifiers (or committee machine) combines multiple classifiers’ results to increase the performance when compared to single classifiers. In this paper, we propose new concepts for combining rules. They are based (1) on uncertainties of the individual classifiers, (2) on combining the result of existing combining rules, (3) on combining local class probabilities with the existing segmentation probabilities at each individual segmentation, and (4) on using uncertainty-based weights for the weighted majority rule. The results show that the proposed local-statistics-aware combining rules can reduce the effect of noise in the individual segmentation result and consequently improve the performance of the final (combined) segmentation. Also, combining existing combining rules and using the proposed uncertainty- based weights can further improve the performance.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Change history

  • 15 March 2018

    The following affiliation, denoted ?d,? should be included for the first author A. Al-Taie: dComputer Science Department, College of Science for Women, Baghdad University, Iraq

  • 15 March 2018

    The following affiliation, denoted ?d,? should be included for the first author A. Al-Taie: dComputer Science Department, College of Science for Women, Baghdad University, Iraq

References

  1. M. N. Ahmed, S. M. Yamany, N. Mohamed, A. A. Farag, and T. Moriarty, “A modified fuzzy c-means algorithm for bias field estimation and segmentation of mri data,” IEEE Trans. Med. Imag. 21 (3), 193–199 (2002).

    Article  Google Scholar 

  2. A. Al-Taie, H. K. Hahn, and L. Linsen, “Uncertainty estimation and visualization in probabilistic segmentation,” Comput. Graph. 39 (1), 48–59 (2014).

    Article  Google Scholar 

  3. X. Artaechevarria, A. Muoz-Barrutia, and C. O. de Solorzano, “Combination strategies in multi-atlas image segmentation: Application to brain MR data,” IEEE Trans. Med. Imag. 28 (8), 1266–1277 (2009).

    Article  Google Scholar 

  4. J. Bezdek, Pattern Recognition with Fuzzy Objective Function Algorithms (New York, Plenum, 1981).

    Book  MATH  Google Scholar 

  5. W. Cai, S. Chen, and D. Zhang, “Fast and robust fuzzy c-means clustering algorithms incorporating local information for image segmentation,” Pattern Recogn. 40 (3), 825–838 (2007).

    Article  MATH  Google Scholar 

  6. S. Chen and D. Zhang, “Robust image segmentation using fcm with spatial constraints based on new kernelinduced distance metric,” IEEE Trans. Syst., Man Cybern.-Part B 34 (4), 1907–1916 (2004).

    Article  Google Scholar 

  7. K. S. Chuang, H. L. Tzeng, S. Chen, J. Wu, and T. J. Chen, “Fuzzy c-means clustering with spatial information for image segmentation,” Comput. Med. Imag. Graph. 30 (1), 9–15 (2006).

    Article  Google Scholar 

  8. T. G. Dietterich, “Ensemble methods in machine learning,” in Proc. 1st Int. Workshop on Multiple Classifier Systems (Springer-Verlag, London, 2000), pp. 1–15.

    Google Scholar 

  9. A. Fred and A. Jain, “Combining multiple clusterings using evidence accumulation,” IEEE Trans. Pattern Anal. Mach. Intellig. 27 (6), 835–850 (2005).

    Article  Google Scholar 

  10. J. Kittler, M. Hatef, R. P. W. Duin, and J. Matas, “On combining classifiers,” IEEE Trans. Pattern Anal. Mach. Intellig. 20 (3), 226–239 (1998).

    Article  Google Scholar 

  11. L. I. Kuncheva, Combining Pattern Classifiers: Methods and Algorithms (Wiley-Intersci., 2004).

    Book  MATH  Google Scholar 

  12. R. Langerak, U. A. van der Heide, A. N. T. J. Kotte, M. A. Viergever, M. van Vulpen, and J. P. W. Pluim, “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).

    Article  Google Scholar 

  13. C. Li, C. Xu, A. Anderson, and J. Gore, “Mri tissue classification and bias field estimation based on coherent local intensity clustering: a unified energy minimization framework,” in Information Processing in Medical Imaging, Lecture Notes in Computer Science (Springer, Berlin, Heidelberg, 2009), Vol. 5636, pp. 288–299.

    Article  Google Scholar 

  14. M. Mignotte, “A label field fusion bayesian model and its penalized maximum rand estimator for image segmentation,” IEEE Trans. Image Processing 19 (6), 1610–1624 (2010).

    Article  MathSciNet  MATH  Google Scholar 

  15. MNI. Brainweb, simulated brain database (1997). http://www.bic.mni.mcgill.ca/brainweb/Accessed Nov. 2012.

  16. N. Mohamed, M. Ahmed, and A. Farag, “Modified fuzzy c-mean in medical image segmentation,” in Proc. IEEE Int. Conf. on Acoustics, Speeeh, and Signal Processing (Piscataway, NI, 1999), Vol. 6, pp. 3429–3432.

    Google Scholar 

  17. M. Paci, L. Nanni, and S. Severi, “An ensemble of classifiers based on different texture descriptors for texture classification,” J. King Saud Univ.–Sci. 25 (3), 235–244 (2013).

    Article  Google Scholar 

  18. T. Rohlfing and C. R. J. Maurer, “Multi-classifier framework for atlas-based image segmentation,” Pattern Recogn. Lett. 26 (13), 2070–2079 (2005).

    Article  Google Scholar 

  19. A. Saad, T. Möller, and G. Hamarneh, “Probexplorer: uncertainty-guided exploration and editing of probabilistic medical image segmentation,” Comput. Graph. Forum 29 (3), 1113–1122 (2010).

    Article  Google Scholar 

  20. A. J. C. Sharkey, Combining Artificial Neural Nets: Ensemble and Modular Multi-Net Systems (Springer-Verlag, New York, 1999).

    Book  MATH  Google Scholar 

  21. L. Szilágyi, Z. Benyo, S. Szilágyi, and H. Adam, “MR brain image segmentation using an enhanced fuzzy cmeans algorithm,” in Proc. 25th IEEE Annu. Int. Conf. Engineering in Medicine and Biology Society (Cancun, 2003), Vol. 1, pp. 724–726.

    Article  Google Scholar 

  22. Y. Tolias and S. Panas, “On applying spatial constraints in fuzzy image clustering using a fuzzy rule-based system,” IEEE Signal Processing Lett. 5 (10), 245–247 (1998).

    Article  Google Scholar 

  23. W. Wang, Y. Zhu, X. Huang, D. P. Lopresti, Z. Xue, L. R. Long, S. Antani, and G. R. Thoma, “A classifier ensemble based on performance level estimation,” in Proc. IEEE ISBI (Boston, MA, 2009), pp. 342–345.

    Google Scholar 

  24. S. K. Warfield, K. H. Zou, and W. M. Wells, “Simultaneous truth and performance level estimation (staple): an algorithm for the validation of image segmentation,” IEEE Trans. Med. Imag. 23, 903–921 (2004).

    Article  Google Scholar 

  25. K. Yuan, L. Wu, Q. Cheng, S. Bao, C. Chen, and H. Zhang, “A novel fuzzy c-means algorithm and its application,” Int. J. Pattern Recogn. Artificial Intellig. 19 (8), 1059–1066 (2005).

    Article  Google Scholar 

  26. D. Zhang and S. Chen, “A novel kernelised fuzzy cmeans algorithm with application in medical image segmentation,” Artificial Intellig. Med. 32 (1), 37–50 (2004).

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Jacobs University, Bremen, Germany

    A. Al-Taie & H. K. Hahn

  2. Fraunhofer MEVIS, Bremen, Germany

    H. K. Hahn & L. Linsen

  3. Westfälischen Wilhelms-Universität, Münster, Germany

    L. Linsen

Authors
  1. A. Al-Taie
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. H. K. Hahn
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. L. Linsen
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to A. Al-Taie.

Additional information

The article is published in the original.

Horst K. Hahn, he studied physics with mathematics, computer science, and physiology as minors at the University of Bayreuth, University Paul Sabatier Toulouse III, and Ruprecht-Karl University Heidelberg. He authored six patents and more than 175 refereed scientific journal and conference publications in the fields of computer-assisted medicine, quantitative medical imaging, computer-aided detection and diagnosis of cancer, neuroimaging, and image-guided therapy. In 2006, he became vice president of the non-profit MeVis Research GmbH and was centrally involved in its transformation into the Fraunhofer Institute for Medical Image Computing MEVIS as of January 2009, where he acts as institute director since Oct. 2012. At the age of 35, he became adjunct professor of medical visualization and four years later professor of medical imaging at Jacobs University Bremen. He is scientific coordinator of several national and European research consortia, and, among other prices and awards, received the Bremer Studienpreis–Special Price Bruker Daltonik for his PhD thesis entitled “Morphological Volumetry–Theory, Concepts, and Application to Quantitative Medical Imaging.” Not least, he organized in July 2011 the prestigious International Conference on Information Processing in Medical Imaging together with Gabor Szekely, ETH Zurich, for the first time in Germany.

Ahmed Al-Taie, he received his B.S. degree in 1997 and M.Sc. degree in 2000 in Computer Science both were from Department of Computer Science, University of Baghdad, Baghdad, Iraq. He joined the Department of Computer Science of College of Science for Women at the University of Baghdad, Iraq, as an assistant teacher in 2005. In fall 2011, he joined the Visualization and Computer Graphics Laboratory (VGCL) group as Ph.D. student in Jacobs University, Bremen, Germany. In 2015, he received his Ph.D. degree. In 2016, he became teacher. His research interests include medical image analysis and processing.

Lars Linsen is a Full Professor (W3) of Computer Science at the Westfälische Wilhelms-Universität Münster, Germany, at the Institute of Computer Science. He is also an Adjunct Professor of Computational Science and Computer Science at the Department of Computer Science and Electrical Engineering of the Jacobs University, Bremen, Germany. He received his academic degrees from the Universität Karlsruhe (TH), Germany, including a diploma (M.Sc.) in Computer Science in 1997 and a Ph.D. in Computer Science in 2001. He spent three years as a post-doctoral researcher and lecturer at the Institute for Data Analysis and Visualization (IDAV) and the Department of Computer Science of the University of California, Davis, U.S.A. He joined the Department of Mathematics and Computer Science of the Ernst-Moritz- Arndt-Universität Greifswald, Germany, as an assistant professor in 2004. In 2006, he joined Jacobs University as an associate professor and became a full professor in 2012. In 2017, he moved to his current affiliation, the Westfälische Wilhelms-Universität Münster, Germany. Lars Linsen’s research interests are mainly in the areas of data visualization or interactive visual data analysis and include certain topics in computer graphics and geometric modeling.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Al-Taie, A., Hahn, H.K. & Linsen, L. Combining rules using local statistics and uncertainty estimates for improved ensemble segmentation. Pattern Recognit. Image Anal. 27, 444–457 (2017). https://doi.org/10.1134/S105466181703004X

Download citation

  • Received:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1134/S105466181703004X

Keywords

Search

Navigation