© 2004 The Royal Microscopical Society Journal of Microscopy, Vol. 216, Pt 1 October 2004, pp. 15– 24 Received 2 February 2004; accepted 19 May 2004 Blackwell Publishing, Ltd. Image calibration in fluorescence microscopy J. M. ZWIER*†, G. J. VAN ROOIJ* 1 , J. W. HOFSTRAAT† & G. J. BRAKENHOFF* *Swammerdam Institute for Life Sciences, Section of Molecular Cytology, University of Amsterdam, Kruislaan 316, 1098 SM Amsterdam, The Netherlands †Institute of Molecular Chemistry, University of Amsterdam, Nieuwe Achtergracht 129, 1018 WS Amsterdam, The Netherlands Key words. Fluorescence microscopy, fluorescence photobleaching, shading correction. Summary A fluorescence image calibration method is presented based on the use of standardized uniformly fluorescing reference layers. It is demonstrated to be effective for the correction of non-uniform imaging characteristics across the image (shad- ing correction) as well as for relating fluorescence intensities between images taken with different microscopes or imaging conditions. The variation of the illumination intensity over the image can be determined on the basis of the uniform bleach- ing characteristics of the layers. This permits correction for the latter and makes bleach-rate-related imaging practical. The significant potential of these layers for calibration in quantita- tive fluorescence microscopy is illustrated with a series of applications. As the illumination and imaging properties of a microscope can be evaluated separately, the methods pre- sented are also valuable for general microscope testing and characterization. Received 2 February 2004; accepted 19 May 2004 Introduction Quantification and calibration of images in fluorescence microscopy is notoriously difficult. Reliable quantification of the fluorescence would permit quantitative comparison of images obtained on different microscopes, or in the same microscope employing different objectives as well as images taken days or weeks apart. For the purpose of this paper the pixellated image P(x,y) – also called in this paper the product distribution – of a fluores- cence microscope can be described as: P(x,y) = I(x,y) · D(x,y) · F(x,y) · t i (s) (1) where I(x,y) is the illumination distribution over the image field of view, D(x,y) the detection efficiency distribution, F(x,y) the fluorescer distribution from pixel to pixel over the speci- men, t i (s) the image exposure time in seconds and x,y the image pixel coordinates. In this paper we address two types of fluorescence calibration: 1 of the fluorescence image intensity. This involves calibra- tion at the level of the product I(x,y) · D(x,y) as needed for shading correction and image comparison, and 2 of the variations in illumination intensity I(x,y) as required for the correction in bleach rate imaging. The key to the approach is the use of fluorescent reference layers for the calibration that are both to a high degree spatially uniform as well as reproducible. In the presented procedure the fluorescence image is calibrated with the help of an image of the reference layer taken under identical imaging condi- tions as the image to be calibrated. The work reported here is a continuation of previously reported work of our group (Ghauharali et al., 1998; Ghau- harali & Brakenhoff, 2000) and is related to the work done by Castleman (1979) and Jericevic et al. (1989). The latter have already shown that with a calibration layer spatial variation of the product of the illumination and detection pathways could be corrected. Ghauharali et al. (1998) have obtained in addi- tion separate illumination distributions by using a mono- exponential function for fitting the observed bleaching of their test layers. In the present work we show that fitting the bleach- ing characteristics using stretched exponential decay kinetics provides much better fits than with a mono-exponential func- tion dependence. The specific value of this method rests in particular with the fact that the reference layers can be manufactured reproducibly and are demonstrated to possess the required uniformity properties. Originally we intended to develop two types of reference layers. One uniformly fluorescing but non-bleaching for calibrating the product distribution P(x,y) and one uniformly bleaching to determine the illumination distribution. Correspondence to: Dr Fred Brakenhoff. Fax: +31 205256271; e-mail: brakenhoff@science.uva.nl 1 Present address: FOM Institute for Plasma physics Rijnhuizen, Postbox 1207, 3430 BE Nieuwegein, The Netherlands.