Biology Reference
In-Depth Information
scanned, such that the entire 2D frame may be illuminated semi-simultaneously.
Fluorescence emitted from each illumination point then returns through the same
aperture of the mask to give a set of confocal points in the image. When the disk is
spun at high speed (typically several thousand revolutions per minute), an appar-
ently continuous image is obtained when viewed through an eyepiece. The resultant
image is traversed with scan lines, but precise synchronization of the illumination
and detection systems virtually eliminates this artifact in well configured systems
(
Wilson et al., 1996
). Thus, this approach allows for very fast scan rates without
significantly compromising the SNR (
Wang et al., 2005
).
As the input illumination power is spread over a much larger area than with
conventional scanning confocal systems, light throughput can be a serious impedi-
ment to successful implementation of spinning disk systems. This was partially
resolved by enhancing the original design by means of a second disk spinning in
sequence with the Nipkow pinhole array disk. This second disk is sited between the
dichroic and the light source and contains a microlens array that maps a miniature
lens to each pinhole (
Tanaami
et al., 2002
), thus improving the illumination
e
ciency by focusing the lightbeam onto the pinhole (
Fig. 5
A). This also reduces
the backscattering of light at the surface of the Nipkow disk, which substantially
increases SNR. The emission detection pathway is not a
Y
ected by this modifica-
tion. Use of specialist cameras and fast versions of the spinning disk head can now
enable imaging rates of up to 2 kHz. However, the drawbacks are that the pinholes
on the spinning disk are inflexible, and the dwell time per pixel is usually very short
(
V
100 ns), which may severely reduce the SNR, although the frequent illumination
of the same pixel as the disk rotates may compensate for this e
V
ect.
IX. Programmable Matrix Microscopy
These instruments are based on the principle of spatially filtering full field
illumination in a defined pattern at high speed, so as to give it a prescribed,
dynamic structure. By using appropriate filtering patterns, the device can simu-
late the optical behavior of confocal scanning microscopes. These systems are
directly comparable to the spinning disk approach in that the illumination device
consists of an array of small apertures that act as both the illumination and
detection pinholes. The principal di
erence, and advantage, is that the elements
of the array are individually addressable, allowing far greater flexibility in exper-
imental design compared to the spinning disk (
Hanley et al., 1999
). As a practical
example, selectively and sequentially illuminating individual cells within the field
is possible with the array system, whereas the Nipkow disk can only operate at
full frame.
Two technologies have been used in implementing practical programmable
matrix systems. The first makes use of a digital micromirror device (DMD), an
array of micrometer-sized mirrors whose angle can be independently controlled
to direct illumination to an ''on'' (confocal) or ''o
V
V
'' (non-confocal) pathway.
