Figure 1. Microscopic imaging as a convolution. (a) Fluorescent beads (0.1 m) were dried onto a coverslip and imaged using a fluorescence microscope, a 63×/1.4 lens and a CCD camera. (b) The same field as in (a) except that the position of the focal plane is shifted 1 m. The distribution of light around each bead is identical, although the total intensity differs depending on the amount of fluorophore in each bead. This image is generated by the convolution of the PSF of the objective lenses with each bead. (c) An axial view of a PSF of a 100×/1.4 lens collected by recording a series of optical sections at 0.1-m intervals through a single 0.1-m bead. The vertical axis of the image is parallel to the optical axis of the microscope.

Figure 2. Normalization of illumination intensity variations. (a) Axial view of three-dimensional data set of a Drosophila embryonic nucleus stained with DAPI, a fluorescent DNA-binding dye. Optical axis and direction of focus change is in the vertical direction. Note the systematic variations in fluorescence intensity at specific optical sections indicated by arrows. (b) As in (a), after restoration, using constrained, iterative deconvolution with an empirical PSF. Deconvolution has enhanced the variations in intensity due to lamp flicker that occurred during data collection. (c) After sampling of total excitation flux during each exposure and normalization of flux differences, systematic intensity variations are corrected. (d) As in (c), after restoration constrained, iterative deconvolution using an empirical PSF. Systematic intensity variations have been removed by measurement and normalization of excitation intensity and the resulting image gives a better representation of the chromosomal structure.