Page 330 - Fundamentals of Light Microscopy and Electronic Imaging

COLOCALIZATION OF TWO SIGNALS USING PSEUDOCOLOR 313

from gray scale to monochrome color is explained in the following way. On luminous
color displays such as monitors and LCD projectors, distinct red, green, and blue pixels
contained in the screen are excited to create the colors of the visual spectrum. For dis-
playing white, the electron guns excite the three color phosphors at maximum, and the
eye perceives the RGB pixel cluster as white. However, for a monochrome red image,
the excitation of green and blue pixels is minimized, so the visual intensity of saturated
red features drops to about a third of the intensity of the same feature shown as white in
a gray-scale image. Therefore, for red monochrome color in an illuminated display, the
brightness and visual dynamic range of the image are greatly reduced.
In color printing, red, green, and blue inks (RGB mode) or cyan, magenta, yellow,
and black inks (CMYK mode) are printed as dots on a sheet of paper. For saturating red
color, inks are applied with the result that only red is reflected from the page to the eye,
while all other wavelengths are absorbed and are not reflected. So, again, for a mono-
chrome red image, saturated features reflect only a fraction of the light that is capable of
being reflected by white in a gray-scale print.
Blue (indigo blue at 400–450 nm) is a particularly difficult color to display (espe-
cially when shown against a black background) owing to the 40- to 100-fold reduced sen-
sitivity of the eye to blue (430 nm) compared to its peak sensitivity to green (550 nm)
(Kingslake, 1965). In color prints, blue chromosomes or nuclei against a dark background
are often disappointing, because little can be perceived or resolved. A much preferred
strategy, particularly for known, defined structures such as chromosomes or nuclei, is to
use cyan (a blue-green color) for the blue color channel. For two-fluorochrome specimens
the image display can be made brighter by choosing green and red colors to which the eye
is very sensitive and avoiding blue altogether. To increase the intensity of the green-red
image, you can change the color palette to show green as yellow-green and red as orange-
red, since both red and green pixels are used to show each of these two colors. This raises
the value of the saturated orange-red pixel by about 30–50% compared to the value for
saturated red alone. Overlap of the two colors still produces yellow, the same as if we had
used only pure red and green. The main point is that the intensity of a saturated feature in
a gray-scale picture is significantly brighter and easier to see than the intensity of the same
saturated feature in a color image. For a three-panel figure showing each of two fluo-
rochromes separately and combined together, an effective strategy is to show the separate
fluorochromes in gray scale, reserving color for just the combined image.

COLOCALIZATION OF TWO SIGNALS USING PSEUDOCOLOR

A common goal in fluorescence microscopy is to determine if two fluorescent signals are
coincident in the same structure. For complex patterns, the use of a combined pseudo-
color display is very useful. Two color channels are selected (red and green) so that
regions of overlap appear yellow. (Other pairings, such as blue and green giving cyan and
blue and red giving magenta, may also be used, but are much less effective because of the
reduced sensitivity of the eye to blue.) In planning for image acquisition and display, it is
best to adopt a standard that is easy to define and understand, namely, combine images
having the same dynamic range for each fluorescent signal. In the case of confocal
microscopy, this practice actually matches the procedure recommended for image acqui-
sition. Once combined, overlapping bright red and green signals give an unambiguous
bright yellow signal. Remember that dark yellow hues, such as colocalized dim objects
and backgrounds with equal red and green contributions, look brown. Optimize the gain