Page 234 - Fundamentals of Light Microscopy and Electronic Imaging

ELECTRONIC ADJUSTMENTS AND CONSIDERATIONS 217

digital converter. The maximum possible DR is the DR of the detector. The DR of the
PMT detector is defined as the ratio of the saturated signal to the detector readout noise,
and is calculated in volts or electrons. The DR of a detector is therefore an inherent fea-
ture of the imaging system. To achieve the imaging potential of the detector, the opera-
tor should try to acquire images that fill its dynamic range, from black (no signal) to
white (saturating signal). This is accomplished by adjusting the amplitude range of the
photon signal using the electronic gain and offset controls of the PMT. The procedures
for adjusting gain and offset are given in this chapter.
Signal-to-noise ratio (S/N) defines the degree of visibility or clarity of an image. It
depends directly on the amplitudes of the object and its background, and on the elec-
tronic noise of the imaging system. For bright images, S/N is determined mainly by the
intensities of the object and background; however, for dim images, the electronic noise
of the imaging system becomes a determining factor. Image brightness at the pinhole is
affected by many variables, including laser power, fluorochrome density in the speci-
men, NA of the objective lens, confocal zoom factor, raster scan rate, the choice of fluo-
rescence filter sets, and other factors. Reduced photon throughput and/or noisy
performance by the detector results in images that lack contrast and look grainy. Gener-
ally, there is little we can do to improve the imaging performance of a detector other
than to provide more light. Typically 50–100 photons/s/pixel are detected from a mod-
erate to bright specimen by CLSM, which corresponds to an S/N ratio of 25 for a sensi-
tive confocal system. This value compares with maximum S/N ratios of 100 to several
hundred for video and digital CCD camera systems. Thus, the image quality of a bright
object in confocal is okay, but not good or excellent. The usual way to improve S/N is to
increase the amount of light by reducing the scan rate or opening the pinhole. Alterna-
tively, a number of individual frames can be averaged, which is called frame averaging
or Kalman averaging, because S/N improves in proportion to the square root of the
number of averaged frames. We return to the topic of improving S/N later. One of the
compromises in obtaining optical slices by confocal imaging is a reduction in photons
contributing to the image and therefore to image quality.
Temporal resolution depends on the raster scan rate and the processing rates of the
detector, the analogue-to-digital converter, and the computer. Frames are typically cap-
tured at a rate of 2 frames/s for a 512 512 pixel image, but rates of 100 frames/s or
higher can be acquired for images of limited size.
In summary, optical performance in the confocal microscope is affected by several
variables, all of which must be evaluated and controlled by the microscopist. It almost
never happens that time, space, and intensity are all resolved optimally for a given com-
bination of specimen and microscope. Generally, it is necessary to make compromises.
For example, to optimize resolution of light intensity to obtain smooth-looking images
with high signal-to-noise ratio, you might need to decrease the spatial and/or temporal
resolution of the image. Recognizing how to balance these parameters requires educa-
tion and experience and is the key to successful confocal imaging.

ELECTRONIC ADJUSTMENTS AND CONSIDERATIONS
FOR CONFOCAL FLUORESCENCE IMAGING

A number of mechanical adjustments, many of which are actuated using the confocal
software, affect the brightness and spatial resolution of the confocal image. These
include adjustments for: