Journal of Mathematical Imaging and Vision

, Volume 37, Issue 3, pp 204–219 | Cite as

Analysis of Bias in the Apparent Correlation Coefficient Between Image Pairs Corrupted by Severe Noise

  • Fredrik BergholmEmail author
  • Jeremy Adler
  • Ingela Parmryd


The correlation coefficient r is a measure of similarity used to compare regions of interest in image pairs. In fluorescence microscopy there is a basic tradeoff between the degree of image noise and the frequency with which images can be acquired and therefore the ability to follow dynamic events. The correlation coefficient r is commonly used in fluorescence microscopy for colocalization measurements, when the relative distributions of two fluorophores are of interest. Unfortunately, r is known to be biased understating the true correlation when noise is present. A better measure of correlation is needed. This article analyses the expected value of r and comes up with a procedure for evaluating the bias of r, expected value formulas. A Taylor series of so-called invariant factors is analyzed in detail. These formulas indicate ways to correct r and thereby obtain a corrected value free from the influence of noise that is on average accurate (unbiased). One possible correction is the attenuated corrected correlation coefficient R, introduced heuristically by Spearman (in Am. J. Psychol. 15:72–101, 1904). An ideal correction formula in terms of expected values is derived. For large samples R tends towards the ideal correction formula and the true noise-free correlation. Correlation measurements using simulation based on the types of noise found in fluorescence microscopy images illustrate both the power of the method and the variance of R. We conclude that the correction formula is valid and is particularly useful for making correct analyses from very noisy datasets.


Correction correlation coefficient Severely degraded images 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Bretzner, L.: Multi-scale feature tracking and motion estimation. Dissertation, Royal Inst. of Tech., Stockholm, Sweden. TRITA-NA-P99/07, Dept. of Numerical Analysis and Computing Science (1999) Google Scholar
  2. 2.
    Pahlavan, K., Uhlin, T., Eklundh, J.-O.: Integrating primary ocular processes. In: Proc. of ECCV 1992, May (1992) Google Scholar
  3. 3.
    Shapiro, L.S., Wang, H., Brady, M.: A corner matching and tracking strategy applied to videophony. Tech. Report OUEL 1933/92, Robotics Research Group, Univ. of Oxford (1992) Google Scholar
  4. 4.
    Francisco, A., Bergholm, F.: On the importance of being asymmetric in stereopsis—or the use of skewed parallel cameras. Int. J. Comput. Vis. 29(3), 181–202 (1998) (Authors in random order) CrossRefGoogle Scholar
  5. 5.
    Manders, E.M.M., Verbeek, F.J., Aten, J.A.: Measurement of co-localization of objects in dual-colour confocal images. J. Microsc. 169, 375–382 (1993) Google Scholar
  6. 6.
    Adler, J., Pagakis, S., Parmryd, I.: Replicate based noise corrected correlation for accurate measurements of colocalization. J. Microsc. 230, 121–133 (2008) CrossRefMathSciNetGoogle Scholar
  7. 7.
    Adler, J., Parmryd, I.: Correlation between sets of data. Patent application in Sweden Nr. 0602596-9. Filed November 29th (2006) Google Scholar
  8. 8.
    Adler, J., Parmryd, I., McKay, I., Pagakis, S.N.: On the quantification and display of colocalisation. Abstract at the Focus on Microscopy meeting, Genua, Italy (2003) Google Scholar
  9. 9.
    Spearman, C.: The proof and measurement of association between two things. Am. J. Psychol. 15, 72–101 (1904) CrossRefGoogle Scholar
  10. 10.
    Spearman, C.: Correlation calculated from faulty data. Br. J. Psychol. 3, 271–295 (1910) Google Scholar
  11. 11.
    Udny, Y.G.: (October 1908) Private letter to Spearman, which later became Appendix E: “Yule’s proof of the correction formula” in Spearman (1910) Google Scholar
  12. 12.
    Udny, Y.G., Kendall, M.G.: An introduction to the Theory of Statistics, 14th edn. Charles Griffin, London (1950). pp. 313–314, 327–328 zbMATHGoogle Scholar
  13. 13.
    Muchinsky, P.M.: The correction for attenuation. Educ. Psychol. Meas. 56(1), 63–75 (1996) CrossRefGoogle Scholar
  14. 14.
    Wilcox, R.: Modern insights about Pearson’s correlation coefficient and least squares regression. Int. J. Sel. Assess. 9(1–2), 195–205 (2001) CrossRefGoogle Scholar
  15. 15.
    Qin, G.Y., Zhu, Z.Y., Fung, W.K.: Robust estimating equations and bias correction of correlation parameters for longitudinal data. Comput. Stat. Data Anal. 52(10), 4745–4753 (2008) zbMATHCrossRefMathSciNetGoogle Scholar
  16. 16.
    Argyros, A., Lourakis, M., Trahanias, P.E., Orphanoudakis, S.: Independent 3D motion detection through robust regression in depth layers. In: Proc. 7th British Mach. Vision Conf. (BMVC 1996), pp. 535–544, Edinburgh, UK, Sept 9–12 (1996) Google Scholar
  17. 17.
    Terzopoulos, D., Witkin, A., Kass, M.: Constraints on deformable models: recovering 3D shape and non-rigid motion. Artif. Intell. 36, 91–123 (1988) zbMATHCrossRefGoogle Scholar
  18. 18.
    Råde, L., Westergren, B.: BETA, Mathematics Handbook, 2nd edn. Chartwell Bratt (1990) Google Scholar
  19. 19.
    Kmenta, J.: Introduction to Econometrics. Macmillan, New York (1971) Google Scholar
  20. 20.
    Brownlee, K.A.: Statistical Theory and Methodology in Science and Engineering. Wiley, New York (1960) zbMATHGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2010

Authors and Affiliations

  • Fredrik Bergholm
    • 1
    Email author
  • Jeremy Adler
    • 2
  • Ingela Parmryd
    • 2
  1. 1.Department of Numerical Analysis and Computer ScienceKTHStockholmSweden
  2. 2.The Wenner-Gren InstituteStockholm UniversityStockholmSweden

Personalised recommendations