PHC1 - Foundations of Confocal Scanned Imaging in Light Micros

PHC1 28/11/2005 6:07 PM Page 1
Foundations of Confocal Scanned Imaging in

Light Microscopy
Shinya Inoué
Seldom has the introduction of a new instrument generated as 1.22l o

instant an excitement among biologists as the laser-scanning con- dmin = (1)
NAobj + NAcond
focal microscope. With the new microscope, one can slice incred-
ibly clean, thin optical sections out of thick fluorescent specimens; where dmin is the minimum spacing in a periodic grating that can
view specimens in planes tilted to, and even running parallel to, just be resolved. dmin is expressed as lateral distance in the speci-
the line of sight; penetrate deep into light-scattering tissues; gain men space; lo is the wavelength of light in vacuum; and NAobj and
impressive three-dimensional (3D) views at very high resolution; NAcond are the numerical apertures of the objective and condenser
obtain differential interference or phase-contrast images in exact lenses, respectively. The NA is the product of the sine of the half-
register with confocal fluorescence images; and improve the pre- angle (a) of the cone of light either acceptable by the objective
cision of microphotometry. lens or emerging from the condenser lens and the refractive
While the instrument that engendered such excitement became indexes (h) of the imbibing medium between the specimen and the
commercially available first in 1987, the optical and electronic objective or condenser lens, respectively.
theory and the technology that led to this sudden emergence had Equation 1 demonstrates that, in addition to the wavelength
been brewing for several decades. The development of this micro- and the NA of the objective lens, the condenser NA also affects
scope stems from several roots, including light microscopy, con- image resolution in the microscope. For objects that are illumi-
focal imaging, video and scanning microscopy, and coherent or nated fully coherently (a condition that pertains when NAcond
laser-illuminated optics (see historic overview in Table 1.1). In this approaches 0, namely when the condenser iris is closed down to a
chapter, I will first discuss some basic principles relating to lateral pinhole), the minimum resolvable lateral spacing increases (i.e.,
and axial resolution as well as depth of field in light microscopy, the resolution decreases) by a factor of 2 compared to the case
highlight some history that lays a foundation to the development when the condenser iris is opened so that NAcond = NAobj. As the
of laser-scanning confocal microscopy, and end with some general condenser iris is opened and NAcond becomes larger, the illumina-
remarks regarding the new microscopes, including a disk-scanning tion becomes progressively less coherent and resolution increases.
confocal system. [Note, however, that laser beams tend to illuminate objects coher-
ently even when the condenser iris is not closed down (see Chapter
5, this volume).]
LIGHT MICROSCOPY Equation 1 describes the relation between NA and resolution
for line-grating objects. A complementary method of defining the
Lateral Resolution1 limit of resolution uses point objects instead of line gratings. The
image, in focus, of an infinitely small luminous object point is
The foundations of modern light microscopy were established a
itself not infinitely small, but is a circular Airy diffraction image
century ago by Ernst Abbe (1873, 1884). He demonstrated how the
with a central bright disk and progressively weaker concentric dark
diffraction of light by the specimen, and by the objective lens,
and bright rings. The radius rAiry of the first dark ring around the
determined image resolution; defined the conditions needed to
central disk of the Airy diffraction image depends on l and the NA
design a lens whose resolution was diffraction limited (rather than
of the objective:
limited by chromatic and spherical aberrations); and established
the role of the objective and condenser numerical apertures (NA) lo
on image resolution (Eq. 1). Thus, rAiry = 0.61 (2)
NAobj
where rAiry is expressed as distance in the specimen plane.

When there exist two points of light separated by a small dis-
tance d in the specimen plane, their diffraction images lie side by
side in the image plane. The images of two equally bright spots
1
For extensive discussions on modern microscope lens design and aberrations are said to be resolved if d is larger or equal to the radius of
and a more rigorous treatment of the optical principles and applications of
light microscopy than is appropriate for this revised chapter, refer to com-
the Airy disk. This is the Rayleigh criterion, and it relies on the
plementary chapters in Handbook of Optics (e.g., Inoué and Oldenbourg, assumption that the two point sources radiate incoherently. If the
1994) and in Video Microscopy, 2nd edition (Inoué and Spring, 1997). two point sources emit light coherently, their amplitude rather than
Shinya Inoué • Marine Biological Laboratory, Woods Hole, Massachusetts 02543
Handbook of Biological Confocal Microscopy, third edition, edited by James B. Pawley, SpringerScience+Business Media, New York, 2006. 1
PHC1 28/11/2005 6:07 PM Page 2
2 Chapter 1 • S. Inoué
TABLE 1.1. Historic Overviewa

Confocal Microscopy Microscopy Video (Microscopy)
Abbe (1873, 1884)a,c Nipkow (1884)
Berek (1927)d
Zworykin (1934)
Zernicke (1935)a,c
Gabor (1948)a
Hopkins (1951)a Flory (1951)
Young and Roberts (1951)
Flying spotc
Linfoot and Wolf (1953)
3D diffraction by annul. apert.a,d
Tolardo di Francia (1955)
Limited fieldb
Nomarski (1955)a
Linfoot and Wolf (1956) Montgomery et al. (1956)
3D diffraction patterna,d Flying spot UVc
Ingelstam (1956)
Resolution and info. theorya
Minsky Patent (1957)
Insighta,b,c,d
Stage scanningf
Kubota and Inoué (1959)a,c
Smith and Osterberg (1961)a
Freed and Engle (1962)
Flying spotc
Harris (1964)a,b
Ellis (1966)
Holomicrographya
Petrán et al. (1968)
Tandem scanningd,g
Davidovits and Egger (1971)
Laser illumination
Lens scanningd
Hellwarth and Christensen (1974)
Second harmonic generationc,d
Hoffman and Gross (1975)
Modulation contrastc
Sheppard and Choudhury (1977)
Theorya,b,d
Sheppard et al. (1978) Ellis (1978)
Stage scanningc,d,e,f Single sideband edge enhancement microscopyc,d
Cremer and Cremer (1978)
Auto-focus stage scanningd
“4-p-point illumination”a,b,d
Brakenhoff et al. (1979) Castleman (1979)
Specimen scana,b,d,e Digital image processinga,c,e
Koester (1980) Quate (1980)
Scanning mirrord Acoustic microscopya,c
Inoué (1981)a,c
Allen et al. (1981a,b)a,c
Fuchs et al. (1982)f
Agard and Sedat (1983)a,c,d
Cox and Sheppard (1983)
Digital recordingd,e
Åslund et al. (1983)
2-mirror laser scanningd
Hamilton et al. (1984)
Differential phasec,d
Wilson and Sheppard (1984)
Extended depth of fieldc,d,e,f
Boyde (1985a) Ellis (1985) Sher and Barry (1985)f
Nipkow typed,e Light scramblerd Fay et al. (1985)a,c,d
Carlsson et al. (1985)
Laser scan, Stacks of confocal imagesd,e
Wijnaendts van Resandt et al. (1985)
xz-viewd,e
PHC1 28/11/2005 6:07 PM Page 3
Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 3
TABLE 1.1. (Continued)

Confocal Microscopy Microscopy Video (Microscopy)
Suzuki-Horikawa (1986) Cox and Sheppard (1986) Inoué (1986)
Video-rate laser scang Overview, How toa,c,e,f
Acousto-optical modulator
No exit pinhole
Xiao and Kino (1987) Castleman (1987)a
Nipkow typed
Amos et al. (1987)c,d,e
McCarthy and Walker (1988) Ellis (1988)
Nipkow typed Scanned aperture phase contrasta,c,d
Denk et al. (1990)
Two photona,c,d,f
Hell and Wichmann (1994)
PSF reduction by stimulated emission depletiona,b,d
Oldenbourg and Mei (1995)
LC-pol systemc,d
Ichihara et al. (1996)
High-throughput spinning diskd, f,g
Conchello et al. (1997)
Aperture scanninga,b,c,d,f
Gustaffson et al. (2000)
Structured illuminationa,b,c,d
Volkmer et al. (2001) Inoué et al. (2001a,b)
Coherent anti-stokes ramanc Centrifuge pol scopec,g
Inoué et al. (2002)
Fluorescence polc
a
Diffraction theory.
b
Superresolution.
c
Contrast modes.
d
Optical sectioning/depth of field.
e
Stereo.
f
3D in objective space.
g
High speed.
their intensity distribution in the image must be considered, and These considerations for resolution assume that the specimen is
resolution generally decreases. The impact of the quality and NA viewed in conventional widefield (WF) microscopy. When the
of the condenser on the lateral resolution of point objects was con- (instantaneous) field of view becomes extremely small, as in con-
sidered by Hopkins and Barham (1950). Their results are similar focal microscopy, the resolution can in fact be greater than when
to, but not strictly identical with, the case of line-grating objects the field of view is not so limited. We shall return to this point
(see Born and Wolf, 1980). later.2
It is important to realize that these resolution criteria apply
only to objective lenses used under conditions in which the image
is free from significant aberrations (see Chapters 6, 7, 8, 9, 11, and
Axial Resolution
22, this volume; Inoué and Oldenbourg, 1994; Chapters 2 and 3 in We now turn to the axial (z-axis) resolution, measured along the
Inoué and Spring, 1997). This implies several things: optical axis of the microscope, that is, perpendicular to the plane
of focus in which the lateral resolution was considered.
• A well-corrected clean objective lens is used within the wave- To define axial resolution, it is customary to use the 3D dif-
lengths of light and diameter of field for which the lens was
fraction image of a point source that is formed near the focal plane.
designed (commonly in conjunction with specific oculars
In the case of lateral resolution, that is, the resolution in the plane
and/or tube lenses).
of focus, the Rayleigh criterion makes use of the infocus diffrac-
• The refractive index, dispersion, and thickness of the coverslip
tion images (the central cross-section of the 3D diffraction pattern)
and immersion media are those specified for the particular
of two point sources and the minimum distance that they can
objective lens.
approach each other laterally, yet still be distinguished as two.
• The correct tube length and ancillary optics are used and the
Similarly, axial resolution can be defined by the minimum distance
optics are axially aligned.
• The full complement of image-forming rays and light waves
leaving all points of the objective-lens aperture converge to the
image plane without obstruction.
2
Note also that one’s ability to determine the location of an object is not deter-
• The condenser aperture is homogeneously and fully
mined by the resolution limit of the system. In fact, the location of an object
illuminated. (diffraction pattern) can be determined under a microscope with precisions
• The condenser is properly focused to produce Köhler that are many times, or even orders of magnitude, greater than the resolution
illumination. limit (e.g., Denk and Webb, 1987; also see Inoué, 1989).
PHC1 28/11/2005 6:07 PM Page 4
that the diffraction images of two points can approach each other microscopy or confocal microscopy with a minute exit pinhole).
along the axis of the microscope, yet still be seen as two. To define The third factor should also disappear once the total magnification
this minimum distance, we use again the diffraction image of an is raised sufficiently, so that the unit diffraction image becomes
infinitely small point object and ask for the location of the first significantly larger than the resolution element of the detector (e.g.,
minimum along the axis of the microscope. Hansen in Inoué, 1986; Castleman, 1987, 1993; Schotten, 1993;
The precise distribution of energy in the image-forming light Inoué and Spring, 1997, Section 12.2).
above and below focus, especially for high NA objective lenses, When the detector can be considered to be infinitely thin and
cannot be deduced by geometric ray tracing but must be derived made up of resolution elements spaced sufficiently (at least 2-fold)
from wave optics. The wave optical studies of Linfoot and Wolf finer than the Airy disk radius, then one need only to consider the
(1956) show that the image of a point source produced by a dif- diffraction-limited depth of field. In that case, the depth of field is
fraction-limited optical system (e.g., a well-designed and properly taken to be
used light microscope) is not only periodic around the point of
focus in the focal plane, but is also periodic above and below the 1
d. = (z - zmin
+ - ) (4)
focal plane along the axis of the microscope. [Such 3D diffraction 4 min
images (including those produced in the presence of lens aberra-
tions) are presented photographically by Cagnet et al. (1962; also that is, one quarter of the distance between the first axial minima
see Chapter 7, this volume, Fig. 7.4). The intensity distribution cal- above (zmin ) and below (zmin ) the central maximum in the 3D Airy
+ -
culated by Linfoot and Wolf for an aberration-free system is repro- pattern converted to distances in specimen space (see Eq. 3; zmin +
duced in Born and Wolf (1980) and also in Inoué and Spring (1997, and zmin correspond to Z1 and -Z1 in Chapter 7, Fig. 7.4, this
-
Fig. 2-30). The 3D pattern of a point source formed by a lens pos- volume).
sessing an annular aperture was calculated by Linfoot and Wolf In conventional fluorescence and darkfield microscopy, the
(1953).] light arising from each image point produces significant intensity
The distance from the center of the 3D diffraction pattern to within a solid cone that reaches a considerable distance above and
the first axial minimum (in object space dimensions) is given by below focus (as seen in the point-spread functions for these modes
of microscopy; e.g., Streibl, 1985; also Chapters 11 and 23, this
2l o h volume). Therefore, fluorescent (or light-scattering) objects that
zmin = (3)
(NAobj)
2
are out of focus produce unwanted light that is collected by the
objective and reduces the contrast of the signal from the region in
where h is the refractive index of the object medium. zmin corre- focus.
sponds to the distance by which we have to raise the microscope For these reasons, the depth of field may be difficult to measure
objective in order to focus the first intensity minimum observed or even to define precisely in conventional fluorescence and dark-
along the axis of the 3D diffraction pattern instead of the central field microscopy. Put another way, one could say that when objects
maximum.3 that are not infinitely thin are observed in conventional fluores-
As with the lateral resolution limit, we can use zmin as a cence or darkfield microscopy, the apparent depth of field is very
measure of the limit of axial resolution of the microscope optics. much greater than the axial resolution.
Note, however, that zmin shrinks inversely proportionally with the The unwanted light that expands the apparent depth of field is
square of the NAobj, in contrast to the lateral resolution limit which exactly what confocal imaging eliminates. Thus, we can view only
shrinks with the first power of the NAobj. Thus, the ratio of axial- those fluorescent and light-scattering objects that lie within the
to-lateral resolution (zmin/rAiry = 3.28 h/NAobj) is substantially larger depth that is given by the axial resolution of the microscope and
than l and is inversely proportional to the NA of the objective lens. attain the desired shallow depth of field.
As mentioned earlier, the lateral resolution of a microscope is
also a function of the size of the field observed at any one instant.
Depth of Field Tolardo di Francia (1955) suggested, and Ingelstam (1956) argued
The depth of field of a microscope is the depth of the image (mea- on the basis of information theory, that one gains lateral resolution
sured along the microscope axis translated into distances in the by a factor of 2 as the field of view becomes vanishingly small.
specimen space) that appears to be sharply in focus at one setting These theoretical considerations set the stage for the development
of the fine-focus adjustment. In brightfield microscopy, this depth of confocal imaging.
should be approximately equal to the axial resolution, at least in
theory. The actual depth of field has been determined experimen-
tally, and the contribution of various factors that affect the mea- CONFOCAL IMAGING
surement have been explored by Berek (1927).
According to Berek, the depth of field is affected by (1) the As a young postdoctoral fellow at Harvard University in 1957,
geometric and diffraction-limited spreading, above and below the Marvin Minsky applied for a patent for a microscope that used a
plane of focus, of the light beam that arose from a single point in stage-scanning confocal optical system. Not only was the concep-
the specimen; (2) the accommodation of the observer’s eye; tion farsighted, but his insight into the potential application and
and (3) the final magnification of the image. The second factor significance of confocal microscopy was nothing short of remark-
becomes irrelevant when the image is not viewed directly through able. [See the delightful article by Minsky (1988) that shows even
the ocular but is instead focused onto a thin detector (as in video greater insight into the significance of confocal imaging than do
the following extracts culled from his patent application.]
In Minsky’s embodiment of the confocal microscope, the con-
3
As discussed later, the distance zmin can be reduced significantly below the ventional microscope condenser is replaced by a lens identical to
classical limit given by Eq. 1.3, for example, by reducing the effective point- the objective lens. The field of illumination is limited by a pinhole,
spread function by special use of two-photon confocal imaging. positioned on the microscope axis. A reduced image of this pinhole
PHC1 28/11/2005 6:07 PM Page 5
is projected onto the specimen by the “condenser.” The field of

view is also restricted by a second (or exit) pinhole in the image
plane placed confocally to the illuminated spot in the specimen and
to the first pinhole (Fig. 1.1). Instead of trans-illuminating the spec-
imen with a separate “condenser” and objective lens, the confocal
microscope could also be used in the epi-illuminating mode,
making a single objective lens serve as both the condenser and the
objective lens (Fig. 1.2).
Using either transmitted or epi-illumination, the specimen is
scanned with a point of light by moving the specimen over short
distances in a raster pattern. (The specimen stage was supported
on two orthogonally vibrating tuning forks driven by electromag-
nets at 60 and 6000 Hz.) The variation in the amount of light, mod-
ulated by the specimen and passing the second pinhole, is captured
by a photoelectric cell. The photoelectric current is amplified and FIGURE 1.2. Optical path in epi-illuminated confocal microscope. The
modulates the beam intensity of a long-persistence cathode-ray entrance pinhole, A, point D in the specimen, S, and exit pinhole, B, are con-
tube (CRT) scanned in synchrony with the tuning forks. As a result, focal points as in Figure 1.1. A partial, or dichromatic, mirror, M1, transmits
the illuminating beam a–b–c and reflects the beam d–e which passed D and
the image of the specimen is displayed on the CRT. The ratio of
was reflected by the mirror, M2, on which the specimen is lying. Only the
scanning distances between the electron beam and the specimen reflected beam that passes point D focuses onto the detector pinhole and reaches
provides image magnification, which is variable and can be very the photocell, P. A single lens, O, replaces the condenser and objective lenses
large. in Figure 1.1. (After Minsky, 1957.)
With this stage-scanning confocal microscope, Minsky says,
light scattered from parts other than the illuminated point on the
specimen is rejected from the optical system (by the exit pinhole)
to an extent never before realized. As pointed out in the patent
application, there are several advantages to such an optical system: • Especially well suited for making quantitative studies of the
optical properties of the specimen
• Reduced blurring of the image from light scattering • An infinite number of aperture planes in the microscope are
• Increased effective resolution potentially available for modulating the aperture with darkfield
• Improved signal-to-noise ratio stops, annuli, phase plates, etc.
• Permits unusually clear examination of thick, light-scattering • Complex contrast effects can be provided with comparatively
objects simple equipment
• xy-scan possible over wide areas of the specimen • Permits use of less complex objective lenses, including those
• Inclusion of a z-scan is possible for long working distance, ultraviolet (UV), or infrared
• Electronic adjustment of magnification imaging, as they need to be corrected only for a single axial
point.
The high-resolution acoustic microscope developed by Quate
and co-workers (Quate, 1980) and the laser disk, video, and audio
recorder/players are object-scanning-type confocal microscopes.
The designers of these instruments take advantage of the fact that
only a single axial point is focused or scanned (see, e.g., Inoué and
Spring, 1997, Sect. 11.10).
IMPACT OF VIDEO
FIGURE 1.1. Optical path in simple confocal microscope. The condenser lens,
C, forms an image of the first pinhole, A, onto a confocal spot, D, in the spec- Nipkow Disk
imen, S. The objective lens, O, forms an image of D onto the second (exit)
Just about the same time that Abbe in Jena laid the foundation for
pinhole, B, which is confocal with D and A. Another point, such as E in the
specimen, would not be in focus with A, so that the illumination would be less. modern light microscopy, a young student in Berlin, Paul Nipkow
In addition, most of the light, g–h, scattered from E would not pass the second (1884), figured out how to convert a two-dimensional (2D) optical
pinhole, B. The light reaching the phototube, P, from E is thus greatly attenu- image into an electrical signal that could be transmitted as a one-
ated compared to that from the confocal point, D. In addition, the exit pinhole dimensional (1D), or serial, time-dependent signal, over a single
can be made small enough to exclude the diffraction rings in the image of D, cable (as in a Morse code). Prior to Nipkow, most attempts at the
so that the resolving power of the microscope is improved. As the specimen is electrical transmission of optical images involved the use of mul-
scanned, the phototube provides a signal of the light passing through sequen- tiple detectors and as many cables.
tial specimen points D1, D2, D3, etc. (not shown). D1, D2, D3, etc., can lie in the Nipkow dissected the image by scanning over it in a raster
focal plane as in conventional microscopy or perpendicular to it, or at any angle
pattern, using a spinning opaque wheel perforated by a series of
defined by the scanning pattern, so that optical sections can be made in or at
rectangular holes. The successive holes, placed a constant angle
angles tilted from the conventional image plane. Because, in the stage-
scanning system, the small scanning spot, D, lies exactly on the axis of the apart around the center of the disk but on constantly decreasing
microscope, the lenses C and O can be considerably less sophisticated than radii (i.e., arranged as an Archimedes spiral), generated the raster-
conventional microscope lenses, which must form images from points some scanning pattern (Fig. 1.3). The brightness of each image element,
distance away from the lens axis. (After Minsky, 1957.) thus scanned by the raster, was picked up by a photocell. The
PHC1 28/11/2005 6:07 PM Page 6
Lewin, 1985). Boyde also provides striking stereoscopic images

obtained with the tandem-scanning confocal microscope (Boyde,
1985b, 1987).
Gordon Kino and co-workers at Stanford University have
designed a confocal microscope using a Nipkow disk in a manner
that differs somewhat from the Petrán̆ type (Xiao and Kino, 1987;
also see Chapter 10, this volume). In the Kino type, the rays illu-
minating the specimen and those scattered by the specimen tra-
verse the same set of pinholes on the spinning Nipkow disk rather
than those that are centrosymmetrical. By using a special low-
reflection Nipkow disk, tilted somewhat to the optic axis of the
microscope, and by employing crossed polarizers and a quarter-
wave birefringent plate to further reduce the spurious reflections
A B from the disk, they are able to use only one side of it, thus allevi-
ating some of the alignment difficulties of the Petrán̆ type.
More recently, spinning-disk confocal units have been vastly
FIGURE 1.3. Nipkow disk. The perforations in the opaque disk, A, which is improved by adding microlenses to the pinholes. As described later
rotating at a constant velocity, scan the image in a raster pattern as shown in (see “Yokogawa Disk-Scanning Confocal System” below), the
B. (After Zworykin and Morton, 1954.) microlens-equipped disk-scanning systems effectively provide
video-rate and faster confocal full-frame imaging and in real color.
While in part depending on older technology, the confocal-
scanning unit (CSU) systems turn out to have certain advantages
not achievable with point-scanning confocal systems.
output of the photocell reflected the brightness of the sequentially
scanned image elements and drove a neon bulb that, viewed
through another (part of the) Nipkow disk, reproduced the desired
Electron-Beam-Scanning Television
picture. While Nipkow’s invention laid the conceptual groundwork for
A similar type of scanner disk, but with multiple, centrosym- television, raster scanning based on a mechanical device was
metric sets of spirally placed holes, was used by Mojmir Petrán̆ simply too inefficient for practical television. Thus, it was not until
and co-workers at Prague and New Haven to develop their epi- five decades after Nipkow, following the advent of vacuum tube
illuminated tandem-scanning confocal microscope (TSM) (Egger and electronic technology, that Zworykin (1934) and his col-
and Petrán̆, 1967; Petrán̆ et al., 1968). In Petrán̆’s microscope, leagues at RCA were able to devise a practical television system.
holes on a portion of the spinning disk placed in front of the light- These workers developed the image iconoscope, an image-storage-
source collector lens are imaged onto the specimen by the objec- type electron-beam-scanning image pickup tube. The image icono-
tive lens. Each point of light reflected or scattered by the specimen scope, coupled with a CRT for picture display, permitted very
is focused by the same objective lens back onto the centrosym- rapid, “inertialess” switching and scanning of the image and
metric portion of the Nipkow disk. The pinholes at this region picture elements. With these major breakthroughs, television not
exclude the light originating from points in the specimen not illu- only became practical for broadcasting but emerged as a tool that
minated by the first set of pinholes, giving rise to confocal opera- could be applied to microscopy (see Inoué and Spring, 1997,
tion (Chapter 10, this volume). Sects. 1.1 and 1.2).
As with Nipkow’s initial attempt at television, this TSM tends An early application of video (the picture portion of television)
to suffer from the low fraction (1%–2%) of light that is transmitted was the flying spot UV microscope of Young and Roberts (1951).
through the source pinholes. Also, very high mechanical precision With this microscope, the specimen remains stationary and single
is required for fabricating the symmetrical Nipkow disk and for object points are scanned serially in a raster pattern by a moving
spinning it exactly on axis. In addition, some of the advantages spot of UV light emitted by the face of a special high-intensity UV-
pointed out by Minsky for the stage-scanning type confocal optics CRT. The optical elements (condenser) of the microscope demag-
are lost because the objective lenses are no longer focusing a single nify this moving spot onto the specimen, which modulates its
axial point of light. brightness. The modulated UV light is then picked up by a photo-
However, for biological applications, the tandem-scanning tube and amplified electronically before being displayed on a
system provides the decided advantage that the specimen remains visible-light CRT scanned in synchrony with the UV-CRT.
stationary. As a result, the speed of the raster scan is not limited Young and Roberts point out that by illuminating only a single
by the mass of the specimen support as it is in stage scanning, and specimen point at a time with a flying-spot microscope, flare is
the scanning system is unlikely to introduce any geometrical dis- reduced and the image becomes a closer rendition of the speci-
tortion. Thus, with a TSM, one can observe objects that reflect or men’s optical properties than that obtained with a non-scanning
scatter light moderately strongly, in real time, either by using a microscope. They also point out that for these same reasons — and
television or photographic camera or by observing the image because a photoelectric detector can provide a sequential, linear
directly through the eyepiece. output of the brightness of each specimen point — quantitative
In addition to Petrán̆ and co-workers, Alan Boyde (1985a) in analysis becomes possible with a flying-spot microscope. In addi-
London took advantage of the good axial discrimination and light- tion, they note that the electronic photodetector raises the sensi-
penetrating capability of the tandem scanning confocal microscope tivity of image capture by perhaps two orders of magnitude
and pioneered its use for viewing biological objects. In particular, compared to photography.
he used it for imaging below the surface of hard tissue such as It should be noted that the flare which would otherwise arise
bone and teeth to visualize the cells and lacunae found there (see from the unilluminated parts of the specimen is significantly
PHC1 28/11/2005 6:07 PM Page 7
reduced with a flying-spot microscope (see Sheppard and 1983; Agard et al., 1989; see also Chapters 23, 24, and 25, this
Choudhury, 1977), even though the exit pinhole used in a confocal volume); 3D imaging including stereoscopy (Brakenhoff et al.,
microscope is not present. Thus, for example, Wilke et al. (1983) 1986, 1989; Inoué and Inoué, 1986; Åslund et al., 1987; Stevens
and Suzuki and Hirokawa (1986) developed laser-scanning flying- et al., 1994; Inoué and Spring, 1997, Sect. 12.7.7); and, finally, the
spot microscopes (coupled with digital image processors) to raise further development of laser-scanning microscopy and confocal
image contrast (at video rate) in fluorescence, differential- microscopy.
interference-contrast (DIC), and brightfield microscopy. Naturally,
the exit pinhole in a confocal system is very much more effective
at excluding unwanted light arising from different layers or por- LASERS AND MICROSCOPY
tions of the specimen not currently illuminated by the source
“pinhole,” but it does so at the cost of reduced image brightness,
lower scanning speed, and increased instrumental complexity and
Holography
price. In 1960, Maiman announced the development of the first operat-
While the flying-spot, or beam-scanning, microscope was ing laser. However, “his initial paper, which would have made his
developed and applied in UV microscopy for about a decade after findings known in a more traditional fashion, was rejected for
its introduction, its further development as an imaging device was publication by the editors of Physical Review Letters — this to
eclipsed for some time by the need and the opportunity to develop their everlasting chagrin.” (For historic accounts including this
automated microscopy for rapid cell sorting and diagnosis. Here, quotation and a comprehensive discussion of the principles
the aim was not the imaging of cell structures as such but rather and application of lasers and holography, see Sects. 14.2 and
the rapid and efficient classification of cells based on their bio- 14.3 in Hecht, 1987; see also Chapter 5, this volume.) Shortly
chemical characteristics, taking advantage of the emerging power thereafter, two types of applications of lasers were sought in
of high-speed digital computers. The size, shape, absorbance, light microscopy. One took advantage of the high degree of monochro-
scattering, or light emission of cells (labeled with specific fluores- maticity and the attendant long coherence length. Coherence
cent markers) was used either to classify the cells by scanning the length is the distance over which the laser waves could be shifted
slide under a microscope or to sort the cells at very high rates as in path and still remain coherent enough to display clear inter-
the cells traversed a monitoring laser beam in a flow cell or a ference phenomena (note that, in fact, this reflects a very high
Coulter-type cell separator. degree of temporal coherence). These characteristics made the
laser an ideal source for holography (Leith and Upatnieks, 1963,
1964).
Impact of Modern Video To explore the use of holography with the microscope, Ellis
Meanwhile, starting in the late 1970s, the introduction of new solid (1966) introduced a conventional light microscope into one of two
state devices, especially large-scale integrated circuits and related beams split from a laser. When this beam was combined with the
technology, led to dramatic improvements in the performance other beam passing outside of the microscope, the two beams could
and availability, and reduction in price, of industrial-grade video be made to interfere in a plane above the ocular. The closely spaced
cameras, video tape recorders, and display devices. Concurrently, interference fringes were recorded on very fine-grained photo-
ever more compact and powerful digital computers and image- graphic film to produce the hologram.
processing systems appeared in rapid succession. These advances What Ellis found was that the coherence length of the laser
led to the birth of modern video microscopy, which in turn brought beam was so long that the hologram constructed as described
about a revitalized interest in the power and use of the light micro- above could be viewed not only to reconstruct an image of the
scope (for reviews see Allen et al., 1981a,b; Inoué, 1981, 1989; specimen being magnified by the microscope, but also to recon-
Allen, 1985; Inoué and Spring, 1997). struct images of the inside of the microscope. Indeed, in the holo-
In brief, dynamic structures in living cells could now be visu- gram one could see the whole optical train and interior of the
alized with a clarity, speed, and resolution never achieved before microscope, starting with the substage condenser assembly, the
in DIC, fluorescence, polarized-light, darkfield, and other modes specimen, the objective lens and its back aperture, the interior of
of microscopy; the growth and shortening of individual molecular the body tube up to the ocular, and even the light shield placed
filaments of tubulin and f-actin, and their gliding motion and inter- above it! This made it possible for Ellis to view the hologram
action with motor molecules, could be followed in real time through appropriately positioned stops, phase plates, etc., and to
directly on the monitor screen; and the changing concentration and generate contrast from the specimen in imaging modes such as
distribution of ions and specific protein molecules tagged with flu- darkfield or oblique illumination, phase contrast, etc., after the
orescent reporter molecules could be followed, moment by hologram itself had been recorded. In other words, the state of the
moment, in physiologically active cells (Chapters 19, 29, and 42, specimen at a given point of time could be reconstructed and
this volume). viewed after the fact in contrast modes different from the one
In addition to its immediate impact on cellular and molecular present when the hologram was recorded.
biology, video microscopy and digital image processing also stim- In principle, holomicrography presents many intriguing possi-
ulated the exploration of other new approaches in light microscopy bilities including 3D imaging. But the very virtue of the long
along several fronts. These include the development of ratio coherence length of the laser beam means that the hologram also
imaging and new reporter dyes for quantitative measurement of registers all the defects and dirt in the microscope. Without laser
local intracellular pH, calcium ion concentration, etc. (Tanasugarn illumination, the optical noise produced by these defects would be
et al., 1984; Bright et al., 1989; Tsien, 1989; Chapters 16, 19, and far out of focus. With a laser illuminating the whole field of view
29, this volume); the computational extraction of pure optical sec- of the microscope, the interference fringes from these out-of-focus
tions from whole-mount specimens in fluorescence microscopy defects intrude into the holographic image of the specimen where
(based on deconvolution of multi-layered images utilizing knowl- they are prominently superimposed. Because of this problem,
edge of the microscope’s point-spread function; Agard and Sedat, holomicrography has so far not been widely used. [However, see
PHC1 28/11/2005 6:07 PM Page 8
Sharnoff et al. (1986), who have figured out how to obtain holomi- Multi-Length Fiber Scrambler
crograms that display only the changes taking place in the speci- None of the mechanical scramblers mentioned above can be used
men (contracting muscle striations) over an interval of time and where speckles have to be removed within extremely brief time
thus eliminate the fixed-patterned optical noise.] periods. For example, in a centrifuge polarizing microscope, the
laser output must be made spatially incoherent within the few
nanoseconds required to freeze the image of the specimen flying
Laser Illumination through the field of the objective lens at speeds up to 100 m/s
Another practical application of lasers in microscopy is its use as (Inoué et al., 2001b). Our solution for reducing the coherence of
an intense, monochromatic light source. Lasers can produce light the laser pulse was to introduce a fiber bundle made up of up to
beams with a very high degree of monochromaticity and polariza- 100 fibers of multiple lengths between two multi-mode single-fiber
tion, implying a high degree of coherence. Some lasers also gen- scramblers. The first multi-mode fiber introduced some phase ran-
erate beams with very high intensity. Thus, an appropriate laser domizing effects, while the multi-length fibers provided a fiber
could serve as a valuable light source in those modes of bundle output whose phase varied depending on the length of the
microscopy where monochromaticity, high intensity, and a high fiber. However, the intensities of the fiber output varied depending
degree of coherence and polarization are important. on their location in the bundle. The final multi-mode fiber made
To use the laser as an effective light source for microscopy, the non-uniform brightness of the bundle output homogeneous, so
three conditions must be satisfied: that the microscope condenser received uniform illumination.
Thus, without using any mechanically moving parts, the phase of
• Both the microscope’s field of view and the condenser aper- the monochromatic laser beam is randomized, and speckles are
ture must appropriately be filled. eliminated from the field image, while the whole condenser aper-
• The coherence length of the laser beam (i.e., the temporal ture if filled uniformly.
coherence) must be reduced to eliminate interference from out-
of-focus defects. Field Scanning
• The coherence at the image plane must be reduced to elimi- The field is scanned by a minute focused spot (the diffraction
nate laser “speckle” and to maximize image resolution. image) of a single-mode laser beam that has been expanded to fill
In fact, these three conditions are not totally independent, but they the condenser aperture (as in a laser-scanning confocal micro-
do specify the conditions that must be met. scope). Thus, the specimen is scanned point by point, and the
One of the following five approaches can be used to fulfill signal light reflected, transmitted, or emitted by the specimen is
these conditions (see also Chapter 6, this volume). collected and focused by the objective lens. This imaging mode
avoids the generation of speckle from laser-illuminated specimens
because speckle arises from the interference between the coherent
Spinning-Disk Scrambler light waves scattered from different parts of a specimen. (This
The laser beam, expanded to fill the desired field, is passed through
optical setup is less effective at removing speckle when a smooth
a spinning ground-glass diffuser placed in front of the beam
reflecting surface is presented slightly away from the plane of
expander lens (Hard et al., 1977). The ground glass diffuses the
focus.)
light so that the condenser aperture is automatically filled.
This fourth approach leads to field-scanning microscopy. A
However, if the ground glass were not moving, small regions of
focused spot of laser light can be made to scan the field as in a
its irregular surface would act as coherent scatterers and the image
flying-spot microscope, or the specimen can be moved and
field would still be filled with laser speckle. Spinning or vibrating
scanned through a fixed focus point. Alternatively, an exit pinhole
the ground glass reduces the temporal coherence of each of the
and beam scanners can be added to generate a laser-scanning con-
coherent scattering points to a period shorter than the integration
focal microscope.
time of the image sensor. Thus, when averaged over the period of
the motion, the field also becomes uniformly illuminated. This Aperture Scanning
approach, while simple to understand, can result in considerable The minute diffraction image of a single-mode laser is focused by
light loss at the diffuser. Also, inhomogeneity of the diffuser’s a beam expander onto an off-axis point on the condenser aperture.
texture can give rise to concentric rings of varying brightnesses The small spot is scanned (made to precess) over the condenser
which traverse the field. aperture in such a way that the field is uniformly illuminated. At
any instant of time, the specimen is illuminated by a tilted colli-
Oscillating-Fiber Scrambler mated beam of light emerging from the condenser and originating
The laser beam is focused onto the entrance end of a single- from the illuminated aperture point. Selected regions of the aper-
stranded multi-mode optical fiber whose output end lies at the focal ture are filled in rapid succession by scanning the spot, so that the
point of a beam-expanding lens. This lens projects an enlarged whole field is illuminated by collimated, coherent beams at suc-
image of the fiber tip to fill the condenser aperture. The fiber, which cessively changing azimuth angles. The rapid scanning of the
is fixed at both ends, is vibrated at some point along its length. The source reduces the temporal coherence of illumination at the object
field and aperture are then uniformly filled with incoherent light plane to less than the response time of the image detector. Never-
with little loss of intensity (Ellis, 1979). If the fiber were not theless, the lateral coherence is maintained for each instantaneous
vibrated, the simple fact that the light beam is transmitted through beam that illuminates the specimen (Ellis, 1988).
the fiber could make the laser beam highly multi-modal. That Ellis has argued the theoretical advantage provided by this fifth
would reduce the lateral coherence of the beam at the aperture approach and has demonstrated its practical attractiveness. With
plane, but the image would still be filled with speckle. Vibration aperture scanning, one gains new degrees of freedom for optical
that reduces the temporal coherence of the beam below the inte- image processing because the aperture function (which controls
gration time of the image sensor integrates out the speckle without the image transfer function of the microscope) can be regulated
loss of light (see also Chapter 6, this volume). dynamically for each point of the aperture. The image resolution
PHC1 28/11/2005 6:07 PM Page 9
and the shallow depth of field that can be achieved with aperture- gradient of the steps could be displayed in a DIC-like image
scanning phase-contrast microscopy is most impressive (see, e.g., (Hamilton and Wilson, 1984). [For the basics of digital image pro-
Inoué and Spring, 1997, Fig. 2-47). cessing, see Castleman (1979), Baxes (1984), Gonzales and Wintz
(1987), Chapter 12 in Inoué and Spring (1997), and Chapter 14,
this volume.]
Laser-Illuminated Confocal Microscopes The integrated circuit chip could also be displayed with con-
During the early 1970s, Egger and co-workers at Yale University trast reflecting the status of the local circuit elements, for example,
developed a laser-illuminated confocal microscope in which the reflecting its temperature or the amount of photo-induced current
objective lens was oscillated in order to scan the beam over the flowing through the circuit, superimposed on the confocal image
specimen. Davidovits and Egger obtained a U.S. patent on this of the chip made with reflected light (Wilson and Sheppard, 1984).
microscope (1972; see review by Egger, 1989). In addition to the Oxford group, the brothers Cremer and
A few years later, Sheppard and Choudhury (1977) provided Cremer (1978) of Heidelberg designed a specimen-scanning laser-
a thorough theoretical analysis on various modes of confocal and illuminated confocal microscope. This epi-fluorescence system
laser-scanning microscopy. The following year, Sheppard et al. was equipped with (1) a circular exit pinhole, in front of the first
(1978) and Wilson et al. (1980) described an epi-illuminating con- PMT, whose diameter was equal to the principle maximum of
focal microscope of the stage-scanning type, equipped with a laser the diffraction pattern; and (2) an annular aperture, in front of a
source and a photomultiplier tube (PMT) as the detector, using a second PMT, whose opening corresponded to the first subsidiary
novel specimen holder. The specimen holder, supported on four maximum of the diffraction pattern. The output of the two PMTs
taut steel wires running parallel to the optical axis, allowed precise was used to provide autofocus as well as displays of surface
z-axis positioning as well as fairly rapid voice-coil-actuated contour and fluorescent intensity distribution.
scanning of the specimen in the xy-plane. Using this instrument, In the 1978 article, the Cremers also discussed the possibility
Sheppard et al. demonstrated the value of the confocal system of laser spot illumination using a “4π-point hologram” that could,
particularly for examining integrated circuit chips. With stage- at least in principle, provide long working distance relative to the
scanning confocal imaging, optical sections and profile images small spot size that could be produced.
could be displayed on a slow-scan monitor over areas very much
larger than can be contained within the field of view of any given
objective lens by conventional microscopy. CONFOCAL LASER-SCANNING MICROSCOPE
These authors capitalized on the fact that the confocal signal
falls off extremely sharply with depth, and the image is therefore In addition to those already mentioned, the pioneering work of the
completely dark for regions of the specimen that are not near the Oxford electrical engineering group was followed in several Euro-
confocal focus plane. For example, with a tilted integrated circuit pean laboratories by Brakenhoff et al. (1979, 1985), Wijnaendts
chip, only the portion of the surface within the shallow depth of van Resandt et al. (1985), and Carlsson et al. (1985). These inves-
field (at any selected z-value) could be displayed, as a strip-shaped tigators respectively developed the stage-scanning confocal
region elongated parallel to the chip’s axis of tilt. Other areas of microscope further, verified the theory of confocal imaging, and
the image were dark and devoid of structure. Conversely, by com- expanded its application into cell biology. I shall defer further dis-
bining all the xy-scan images made during a slow z-scan, they cussions on these important contributions to authors of other chap-
could produce a final “extended focus” image of the whole tilted ters in this volume. In the meantime, video microscopy and digital
surface, which demonstrated maximum spatial resolution on all image processing were also advancing at a rapid rate.
features throughout the focus range (Wilson and Sheppard, 1984; These circumstances culminated in the development of the
Wilson, 1985) (Chapter 22, this volume). confocal laser-scanning microscope (CLSM, Figs. 1.4, 1.5; Åslund
This could be done even when the specimen surface was not et al., 1983, 1987) and publication of its biological application by
a single tilted plane but was wavy or consisted of complex sur- Carlsson et al. (1985), Amos et al. (1987), and White et al. (1987).
faces. In their monograph Scanning Optical Microscopy, Wilson The publications were followed shortly by introduction of laser-
and Sheppard (1984) show shallow optical sections of insect anten- scanning confocal microscopes to the market by Sarastro, Bio-
nae shining on a dark background. They also show stereo-pair Rad, Olympus, Zeiss, and Leitz. It was White, Amos, and Fordham
images of the same object consisting of two “extended focus” of the Cambridge group that first enraptured the world’s biologi-
images made by focusing along two focal axes that were tilted by cal community with their exquisite and convincing illustrations of
several degrees relative to the optical axis. Extended-focus images the power of the CLSM. Here at last was a microscope that could
demonstrate that the confocal system can either decrease or generate clear, thin optical sectioned images, totally free of out-of-
increase the effective depth of field without loss of resolution. focus fluorescence, from whole embryos or cells and at NAs as
As described in the final section of this article, the lateral res- high as 1.4. Not only could one obtain such remarkable optical-
olution that is practically attainable can be improved by using con- sectioned fluorescence images in a matter of seconds, but x-z sec-
focal optics. In addition, the removal of the extraneous light tions (providing views at right angles to the normal direction of
contributed by out-of-focus objects dramatically improves the con- observation) could also be captured and rapidly displayed on the
trast and gives rise to a brilliantly sharp image. monitor. A series of optical sections (stored in the memory of the
Sheppard et al. also managed to display different regions on built-in or add-on digital image processor) could be converted into
the surface of an integrated circuit chip with varying intensity or 3D images or displayed as stereo pairs. The confocal fluorescent
pseudocolor corresponding to the height of the region. This is pos- optical sections could also be displayed side by side with non-
sible because the amount of light reflected by an (untilted) step on confocal brightfield or phase-contrast images, acquired concur-
the surface of the chip and passing the second pinhole varies with rently using the transmitted portion of the scanning laser beam.
the distance of the reflecting surface from the focal plane. The These images could also be displayed superimposed on top of each
authors also showed that, by processing the photoelectric signal other, for example, with each image coded in different pseudo-
electronically, the edges of the steps alone could be outlined or the color, but unlike similar image pairs produced by conventional
PHC1 28/11/2005 6:07 PM Page 10
FIGURE 1.4. Schematic of laser-scanning confocal micro-

scope. (From Åslund et al., 1987.)
microscopes, the two images were in exact register and showed no full cones of intense excitation light, converging and diverging
parallax as each was generated by the same scanning spot. from the illuminated spot. Thus, with conventional confocal
Most of the laser-scanning systems discussed in this section microscopy, biological specimens tend to suffer from photon-
employed epi-illumination using some form of mechanical scan- induced damage and rapid bleaching of fluorescence, while the
ning devices. They could not readily be applied to confocal fraction of the short wavelength excitation beam that reaches the
imaging of transmitted light, for example, for high-extinction focal plane is reduced by absorption in the intervening material.
polarization or DIC microscopy. Nevertheless, Goldstein et al. Many of these shortcomings are circumvented by two- and multi-
(1990) developed a system using an Image Dissector Tube which, photon microscopy.
in principle, should be able to provide confocal imaging in the By focusing a pulse of very intense laser beam with twice the
trans-illumination mode. Such an approach may eventually lead to wavelength (half the frequency) of the standard short wavelength
workable transmission laser-scanning confocal microscopes with excitation beam, and within a period shorter than the fluorescence
multiple contrast modes. decay time of the fluorophore, the coherently interfering photons
can excite molecules at half the wavelength of the long wavelength
laser, and do so selectively in the focused spot. In other words, the
TWO- AND MULTI-PHOTON MICROSCOPY output of an intense near-infrared (IR) laser induces fluorescence
in a blue or UV excitable fluorophore at, and only at, the focused
As noted, conventional point-scanning confocal microscopes dra- spot where the coherent electromagnetic field strength is so high
matically reduce the contribution of fluorescence from out-of- (within the required brief period) that it acts nonlinearly to excite
focus regions of the specimen. Nevertheless, regions of the the chromophores at twice the frequency of the IR field. The flu-
specimen above and below the focal plane are exposed to the orophores in the cone of the illuminating light above and below
focus do not experience the two-photon effect and, therefore, are
not excited or damaged. Additionally, in contrast to conventional
confocal microscopy, two-photon laser scanning systems do not
require an exit pinhole or an image-forming objective lens. This is
because the minute two-photon excited fluorescent spot is totally
isolated in space and is free of “parasitic” fluorescence in the xy-
plane as well as along the z-axis. Therefore, the fluorescence emis-
sion needs only to be collected by an efficient photodetector as the
excitation spot is being scanned (see Chapter 28, this volume).
Thus, compared to conventional confocal microscopy, two-
photon microscopy permits confocal imaging of planes much
deeper in the tissue, and with considerably higher light-gathering
efficiency, as well as with less fluorescence bleaching and speci-
men damage outside of the focal plane. Denk et al. (1990) and
Squirrel et al. (1999) have made extended time-lapse recordings
of dividing tissue-cultured cells and mammalian embryos in two-
photon microscopy.
In another interesting and ingenious application of two-photon
microscopy, the fluorescence excitation volume (point-spread
function) has been reduced considerably below that defined by
wave optics by use of two partially overlapping excitation
FIGURE 1.5. Depth discrimination in a laser-scanning confocal fluorescent volumes. The first excites fluorescence in the standard two-photon
microscope. Compare with Figure 1.2. (Courtesy of Dr. H. Kapitza, Carl Zeiss, volume, while the second volume, concurrently generated by a
Oberkochen.) somewhat longer wavelength, quenches the fluorescence (by stim-
PHC1 28/11/2005 6:07 PM Page 11
ulated emission depletion) in the zone where the two volumes speed provided by the Nipkow disk system may be indispensable.
overlap each other. A phase plate in the depletion beam path has For example, at the ~10,000¥ magnification needed for clear visu-
the effect that this beam is almost the inverse of the Airy disk and alization, the Brownian motion of microtubules (even those many
has a null at the focus. Thus, the volume of region actually fluo- micrometers long) is so great that an image acquisition time of
rescing is carved smaller than the standard two-photon excitation >0.1s blurs the image beyond use.
volume, and, in fact, Hell and Wichmann (1994) report having As discussed earlier, the downside of the classical Nipkow
reduced the height of the point-spread function (PSF) by as much disk-type system is that the efficiency of light transmission is low,
as a factor of five (see also Klar et al., 2000, Chapter 31, this light reflected by the spinning disk reduces image contrast, and the
volume). image may suffer from intrusive scan lines. Also, observation is
usually by direct viewing through the ocular, or via some photo-
graphic or video imaging device, rather than using a PMT. While
IS LASER-SCANNING CONFOCAL video imaging does have its own advantages, video sensors other
MICROSCOPY A CURE-ALL? than cooled charge-coupled devices (CCDs) and special return-
beam-type pickup tubes operate over a limited dynamic range.
With the impressively thin and clean optical sections that are Conventional video pickup tubes seldom respond linearly over a
obtainable, and the x-z sections and stereoscopic images that can range of >100 : 1 (more commonly somewhat less; see Inoué and
neatly be displayed or reconstructed, one can be tempted to treat Spring, 1997), and they have relatively high measurement noise.
the CLSM as a cure-all. One may even think of the instrument as By contrast, a PMT can have a dynamic range of ≥106. When
the single microscope that should be used for all modern cell exceedingly weak signals need to be detected from among strong
biology or embryology. How valid is such a statement and what, signals, or when image photometry demands dynamic range and
in fact, are the limitations of the current instruments beyond their precision beyond those attainable with standard video cameras, an
high costs? imaging system using a cooled CCD or a PMT detector may be
The fundamental limits of confocal imaging will be covered required. Modern stage-scanning- and laser-scanning-type con-
in the next chapter. Here I will comment on three topics: the speed focal microscopes use such detectors (Chapter 12, this volume).
of image or data acquisition, comparison with the depth of field Nevertheless, for some applications improved versions of the
in phase-dependent imaging, and some optical and mechanical Nipkow-disk-type confocal instruments may provide optical sec-
factors affecting confocal microscopy. tions with better signal and image quality than with CLSMs as dis-
cussed below under “Yokogawa Disk-Scanning Confocal System.”
The frame-scanning rate of the CLSM falls somewhere
Speed of Image or Data Acquisition between that of the stage- and tandem-scanning types, normally
Several factors affect the time needed to acquire a usable image about 1 to 2 s/frames. This rate is the minimum time required by
with a confocal microscope. These include (1) the type of confo- the mirror galvanometers (that are used to scan the illuminating
cal system used; (2) the optical magnification and numerical aper- and return beams) to produce an image of, say, 512 ¥ 768 picture
ture of the system; (3) the desired area covered; (4) required quality elements. This limitation in scanning speed relates to the absolute
of the image (e.g., lateral and axial resolution, levels of image gray time required to scan along the fastest axis (usually the x, or hor-
scale, degree of freedom from graininess); and (5) the amount of izontal, scan). The scanning speed cannot be increased without
light reaching the sensor. Here we will survey a few general points affecting image resolution or confocal discrimination (Chapters 3,
relating to the choice of instruments, specifically as applied to 21, and 25, this volume).
biology. The x-scanning speed can be increased by using a resonance
Among the different confocal systems, the stage-scanning type galvanometer, a spinning mirror, or an acousto-optical modulator
requires the longest time (~10 s) to acquire a single image because instead of the mirror galvanometers (Chapters 3, 9, and 29, this
the specimen support has to be translated (vibrated) very precisely. volume). However, doing so may reduce both scan flexibility (i.e.,
Biological specimens are often bathed in a liquid medium, and for no optical “zoom” magnification) and inefficient use of the duty
these, any movement presents a problem. Even if the specimen cycle. Furthermore, in a scanning confocal system used for fluo-
chamber is completely sealed and the gas phase excluded to min- rescence microscopy, one cannot use the same acousto-optical
imize the inertial effects of stage scanning, specimen motion still device (or other diffraction-based electro-optical modulator) to
can occur during stage scanning. The alternative lens-scanning both scan the exciting beam and de-scan the emitted beam because
system can encounter worse problems when oil-immersion lenses the modulator would deviate the two beams by different amounts
are used. Very often structures in biological specimens are moving based on their l.
or changing dynamically at rates incompatible with very slow scan Of even greater importance, the image captured by a CLSM
rates. Thus, despite the many virtues of the stage-scanning system in a single, 1- to 2-s scan time is commonly too noisy because the
recognized by Minsky (1957) and by Wilson and Sheppard (1984), image-forming signal is simply not made up of enough photons.
there is little chance that the stage-scanning microscope will be The image generally must be integrated electronically over several
widely used in biology. An exception might be for large-area 3D frame times to reduce the noise, just as when one is using a high-
scanning of fixed and permanently mounted specimens. Such spec- sensitivity video camera. Thus, with a CLSM, it often requires
imens require, or can take advantage of, those virtues of the stage- several, or many, seconds to acquire a well-resolved, high-quality
scanning system that cannot be duplicated by other confocal fluorescence image.
designs. If, in an attempt to reduce the number of frames that must be
In the Petrán̆-type TSM or the Kino-type confocal microscope, integrated, one tries to increase the signal reaching the PMT by
the disk can be spun rapidly enough to provide images at video raising the source brightness, by opening up the exit pinhole, or
rate (30 frames/s). When speed of image acquisition is of para- by increasing the concentration of fluorochrome, each alteration
mount importance, as in the study of moving cells, living cells at introduces new problems of its own. In fact, in CLSMs used for
high magnification, or microtubules growing in vitro, the type of fluorescence imaging, if anything, one wants to reduce the light
PHC1 28/11/2005 6:07 PM Page 12
reaching the specimen in order to avoid saturation of the fluo-

rophores, significant bleaching, and other excitation-induced
damage. There is almost an indeterminacy principle operating
here: One simply cannot simultaneously achieve high temporal
resolution, high spatial resolution, large pixel numbers, and a wide
gray scale simultaneously. This speed limitation must be seen as a
disadvantage of the CLSM.
As already discussed, two-photon confocal fluorescence micro-
scopy (Chapter 28, this volume) is a promising new approach
that may reduce the effect of some of these limitations in addition
to providing excellent lateral and axial resolution. However,
because the time between pulses is long (10–12 s) compared to
the fluorescence time of organic dyes, it only produces signal 10%
to 20% of the time. This low-duty cycle exacerbates the data rate
limit.
While the sampling rate for obtaining whole images with the
CLSM is limited, this does not imply that the temporal resolution
of the detector system is inherently low. For example, one can
measure relatively high-speed events with the CLSM, if one
decides to sacrifice pixel numbers by reducing the size of the
scanned area or even by using a single, or a few, line scan(s). In
addition to the high temporal resolution, the bleaching of diffusible FIGURE 1.6. Schematic of optics in the CSU-10. The expanded and colli-
fluorochromes and photodynamic damage to the cell are reported mated laser beam illuminates the active portion of the upper Nipkow disk con-
taining some 20,000 microlenses. Each microlens focuses the laser beam
to be significantly reduced when the scan is restricted to a single
(through the dichromatic filter cube) onto its corresponding pinhole, thus sig-
line (Chapter 19, this volume). nificantly raising the fraction of the illuminating beam that is transmitted by
Another alternative for gaining speed is to use a slit instead of the main Nipkow disk containing the pinhole array. From the pinholes, the
a pinhole for confocal scanning. This approach, although some- beams progress down to fill the aperture of the objective lens. The objective
what less effective than confocal imaging with small round pin- lens generates a reduced image of the pinholes in the focal plane of the
holes, is surprisingly effective in suppressing the contribution specimen.
of out-of-focus features. Several manufacturers have produced Fluorescence given off by the illuminated points in the specimen is cap-
laser-illuminated, slit-scanning confocal microscopes that provide tured by the objective lens and focused back onto the Nipkow disk containing
video-rate or direct-view imaging systems that are quite easy to the pinhole array. Each pinhole now acts as its own confocal exit pinhole and
operate, at a fraction of the price of the normal CLSM. However, eliminates fluorescence from out-of-focus regions, thus selectively transmitting
fluorescence that originated form the specimen points illuminated by that par-
the rapid bleaching of fluorescent dyes encountered with the slit-
ticular pinhole. (However, for specimens with fluorescence distributed over
scanning system has been a disappointment for those hoping to large depths, some out-of-focus fluorescence can leak through adjacent pin-
gain confocal scanning speed for studies on living cells. holes in multiple pinhole systems such as the CSU-10; see text.) The rays trans-
mitted by the exit pinholes are deflected by the dichromatic beam splitter,
Yokogawa Disk-Scanning Confocal System located between the two Nipkow disks, and proceed to the image plane. (Figure
courtesy of Yokogawa Electric Corporation.)
A new confocal disk-scanning unit (CSU-10 and CSU-21)
designed by Yokogawa Electric Corporation provides video-rate
and faster confocal imaging with several advantages while over-
coming the two major factors that had limited earlier TSM tive as, or the confocal stringency does not match that of, a point-
systems. The new system uses two Nipkow-type disks located one scanning confocal system. Nevertheless, the Yokogawa system
above the other with precisely aligned perforations. In place of pin- provides very effective focal discrimination and capability for pro-
holes, the first disk contains some 10,000 microlenses, each of viding striking optical sections in real time, either for direct obser-
which focuses the collimated laser beam onto a corresponding vation through the eyepiece or captured on a video or photographic
pinhole on the second disk. The microlenses increase the through- camera or a CCD. The residual out-of-focus contribution can be
put of excitation laser from a scant 1% to 2% of conventional rapidly and effectively reduced by “unsharp masking” or “neigh-
Nipkow disks to nearly 50%. At the same time, a dichromatic filter borhood deconvolution” digital processing. (Several examples of
cube is placed between the two disks, so that light reflected or scat- the dynamic cellular changes captured with the CSU-10, as well
tered from the initial disk no longer contributes unwanted back- as the effectiveness of postprocessing, are illustrated in Inoué and
ground to the fluorescence signal received by the detector (Fig. Inoué, 2002. See also Chapter 10, this volume.)
1.6). These confocal scanning units can be attached to any upright In addition to the effective, real-time and faster-than-video-rate
or inverted research-grade light microscope. The 1000 or so pin- confocal fluorescence imaging in real color (which can be viewed
holes that scan the specimen in parallel at any instant of time are superimposed with the brightfield or DIC image of the specimen),
arranged in a unique geometrical pattern. The unique pattern several observers have been impressed by the significantly slower
reduces image streaking (found with conventional Nipkow disks) fluorescence bleaching rate and much longer survival time for
and provides uniform illumination of the whole field of view living cells observed with the CSU-10 (coupled to low-noise CCD
(Inoué and Inoué, 2002). or video cameras) compared to imaging of the same objects with
With any multiple-pinhole- (or slit-) scanning system, some point-scanning confocal systems. For example, CSU-10 imaging
light originating from outside the focal plane is transmitted through was found indispensable for capturing the dynamic growth,
“neighboring” pinholes, so that focal discrimination is not as effec- motion, and gliding of GFP-expressing microtubules in yeast cells
PHC1 28/11/2005 6:07 PM Page 13
as well as speckle images of tubulin flux in pTk-1 tissue cells and the commercially available phase annulus was rather small, and
in the thick spindles undergoing mitosis in Xenopus egg extracts out-of-focus regions intruded obtrusively into the image. With
(Waterman-Storer and Salmon, 1997; Grego et al., 2001; Tran et Ellis’ aperture-scanning phase-contrast microscope, the illuminat-
al., 2001; Maddox et al., 2002 ). The reasons for low-fluorescence ing rays, and the correspondingly minute phase absorber spot, scan
bleaching and extended cell survival are discussed in Inoué and the outermost rim of the objective lens aperture in synchrony.
Inoué (2002 and in Chapter 38, this volume.) Therefore, essentially the full NA of the objective lens is available
The advantage of the real-time, direct-view confocal system to transmit the waves diffracted by the specimen. Under these con-
extends beyond capturing sharp images of moving or dynamically ditions, the z-axis resolution of the optical section in phase-
changing objects, whose images would be blurred or distorted by contrast appears to be even higher than that of the two other
the slow frame-capture rate of conventional CLSMs. For example, contrast modes shown here (Ellis, 1988; Inoué, 1994).]
with a CLSM it is difficult to visualize, or even to find, minute flu- For polarization microscopy of specimens with low retar-
orescent objects that are sparsely distributed in three dimensions. dances, the LC-Pol scope system devised by Oldenbourg and Mei
With an image intensifier CCD camera coupled to a direct-view (1995; see also Oldenbourg, 1996) also provides effective optical
system, the signal from such sparsely distributed objects is readily sectioning. The LC-Pol scope generates an image (retardance map)
found in real time as one focuses through the specimen. whose pixel brightness is proportional to the retardance of the
specimen at each pixel, independent of the specimen’s azimuth ori-
entation, while the algorithm used to compute the retardance map
Depth of Field in Phase-Dependent Imaging also reduces the polarization aberrations introduced by the optics
The z-axis resolution measured in epi-fluorescence imaging with (that otherwise degrade the image; Shribak et al., 2002). Thus,
a confocal laser scanning microscope is reported to be 1.5 mm with with the LC-Pol scope system, objective lenses with NA as high
an NA 0.75 objective lens (Cox and Sheppard, 1993) and 0.48 mm as 1.4 can be used at their full aperture to detect retardances as low
with an NA 1.3 objective lens (Hell et al., 1993) at a wavelength as 0.03 nm. The use of the high NA lenses at full aperture then pro-
of 514 nm. Kino reports a depth of field of 0.35 mm for NA 1.4 con- vides the shallow depth of field (of less than 1-mm thickness) as
focal optics, when imaging point-like reflecting objects. These illustrated in Figure 1.8. In fact, the LC-Pol scope can individu-
numbers are in good agreement with Eqs. 2 and 3, and the height ally resolve two flagellar axonemes that cross each other and
of the 3D diffraction pattern of a point object discussed earlier. are separated by no more than their diameter of about 0.2 mm
In addition, Stephan Hell, as described above, has achieved even (Oldenbourg et al., 1998).
shallower field depths by superimposing two 3D diffraction spots We do not yet quite understand why the depth of field of the
of differing wavelengths in stimulated depletion point- non-confocal phase-dependent images should be so thin. It may
scanning confocal microscopy. well be that contrast generation in phase-dependent imaging
How do these shallow depth of fields attainable with a confo- involves partial coherence even at very high NAs, and that an
cal microscope compare with those obtainable in the absence of effect similar to the one proposed elsewhere for half-wave masks
confocal imaging? While I could come up with no hard numbers (Inoué, 1989) is giving us increased lateral as well as axial reso-
for fluorescence microscopy without confocal imaging (except lution. Whatever the theoretical explanation turns out to be, our
where 3D deconvolution is employed, see Chapters 23, 24, and 25, observations show that for phase-dependent imaging of relatively
this volume), it is well known that the fluorescence from out-of- transparent objects, even in the absence of confocal optics, optical
focus objects substantially blurs the in-focus image. On the other sections can be obtained (at video rate) that appear to be some-
hand, for contrast generated by phase-dependent methods such as what thinner than for fluorescence imaging in the presence of con-
phase-contrast, DIC, and polarized-light microscopy, Gordon Ellis focal optics. Moreover, they perform this function without
and I have obtained data that show remarkably thin optical sec- requiring that energy be deposited in the specimen i.e. without
tions in the absence of confocal imaging. producing photodamage.
Thus, using a 100¥ NA 1.4 Nikon PlanApo objective lens,
combined with an NA 1.35 rectified condenser whose full aperture
was uniformly illuminated through a light scrambler with 546-nm OTHER OPTICAL AND MECHANICAL FACTORS
light from a 100-W high-pressure Hg arc source (as described in AFFECTING CONFOCAL MICROSCOPY
Ellis, 1985, and in Inoué, 1986, Appendix 3), I obtained depth of
fields of ca. 0.2, 0.25, and 0.15 mm, respectively, for phase-
contrast, rectified DIC, and rectified polarized-light microscopy.
Lens Aberration
These values were obtained by examining video images of With stage- or object-scanning confocal microscopes, we saw
surface ridges on a tilted portion of a human buccal epithelial cell. earlier that high NA lenses with simplified design and long
The video signal was contrast enhanced digitally but without working distances could be used because the confocal image points
spatial filtration. The change in image detail that appeared with (source pinhole, illuminated specimen point, and detector pinhole)
each 0.2-mm shift of focus (brought about by incrementing a cal- all lie exactly on the optical axis of the microscope. This same
ibrated stepper motor) was inspected in the image and enlarged principle is now used widely in the design of optical disk
to ~10,000¥ on a high-resolution video monitor. As shown in recorder/players.
Figure 1.7, the fine ridges on the cell surface are not contiguous In contrast, with TSM and CLSM sharp images of the source
in the succeeding images stepped 0.2 mm apart in the polarized- “pinhole(s)” must be focused over a relatively large area away from
light and phase-contrast images, but they are just contiguous in the the lens axis. In addition, the objective lens and the scanner must
DIC images. From these observations, the depth of field in the rec- bring images of the illuminated spot(s) and the source pinhole(s)
tified polarized-light image is estimated to be somewhat below, into exact register with the exit pinhole(s), and for fluorescence
and the DIC image just above, the 0.2-mm step height. [The phase- microscopy, do so at different wavelengths. Thus, for these systems
contrast images here should not be compared literally with the to function efficiently, the microscope objective lens has to be
images in the two other contrast modes because the diameter of exceptionally well corrected. The field must be flat over an
PHC1 28/11/2005 6:07 PM Page 14
FIGURE 1.7. Optical sections of surface ridges on an oral epithelial cell. These ultrathin optical sections were obtained without confocal imaging in phase-
contrast (left), rectified DIC (middle), and rectified polarized-light microscopy (right). The focus planes for the successive frames in each contrast mode were
incremented 0.2 mm. Scale bar 10 mm. (See text and original article for details. From Inoué, 1988.)
appreciable area, axial and off-axis aberrations must be corrected design, a series of excellent-quality, high-NA, PlanApo, and high-
over the field used, and lateral and longitudinal chromatic aberra- UV-transmitting lenses have appeared from all four major micro-
tions must be well corrected for both the emission and illuminating scope manufacturers (Leitz, Nikon, Olympus, Zeiss) during the
wavelengths. As far as is possible, the aberrations should be cor- past decade.
rected within the objective lens without the need to use a compli- Even with excellent lenses, however, the image loses its sharp-
mentary ocular. [For details of these subjects and design of modern ness when one focuses into a transparent, live, or wet specimen by
lenses to overcome the aberrations, see Inoué and Oldenbourg more than a few micrometers from the inside surface of the
(1994), Shimizu and Takenaka (1994), and Chapter 7, this volume]. coverslip. The problem here is that oil-immersion microscope
Finally, the lens and other optical components must have good objectives are designed to be used under rather stringent optical
transmission over the needed wavelength range. conditions, namely homogeneous immersion of everything,
These combined conditions place a strenuous requirement on including the specimen itself, in a medium of h = 1.52. When such
the design of the objective lens. Fortunately, with the availability a lens is used on live or wet specimens immersed in water or phys-
of modern glass stocks and high-speed computer-optimized iological saline solution, even with the coverslip properly oil con-
PHC1 28/11/2005 6:07 PM Page 15
FIGURE 1.8. Optical sections of meiosis-I metaphase spindle in live spermatocyte of a crane fly Nephrotoma sturalis observed with the LC-Pol scope. In focus
are: (A) the upper kinetochores (Ks) and their two K-fibers for the left bivalent chromosome; (B) upper Ks and fibers for the middle and right bivalents; (C)
lower Ks and fibers for the middle and right bivalents. The birefringence retardation of the K-fibers, made up of a dense bundle of microtubules, is ca. 1 nm
greater than that due to the background array of spindle microtubules. The effect of optical sectioning is more obvious in the mitochondrial threads (which are
much thinner than the K-fibers and free of background birefringence) that surround the spindle. Imaged with 546-nm illumination in LC-Pol scope with Nikon
60¥/1.4 NA “DIC” PlanApo objective lens combined with condenser NA at 1.0. The fine focus control was shifted 1.2 mm between (A) and (B), and 0.9 mm
between (B) and (C). White = 3.0 nm retardance. Scale bar = 5 mm.
tacted to the objective lens, aberrations are no longer properly cor- water between specimen and coverslip. Adjustment of the collar,
rected once the image-forming rays traverse a significant distance as in dry- or variable-immersion objective lenses, compensates for
in the h = 1.33 aqueous medium. Doing so distorts the unit dif- the relative thickness of layers having higher or lower h. With the
fraction image and alters the shape (and even the location of) the increasing use of electronic and electro-mechanical controls in
point-spread function, and does so to varying degrees as one confocal and conventional microscopes, it is now possible to
focuses to different depths into the aqueous medium.4 This impor- design a superior high-NA lens with an auto-compensating cor-
tant topic is discussed in detail in Chapters 7 and 20 (this volume) rection device (possibly built outside of the objective lens) that is
and in Inoué and Spring (1997, Sect. 2.5). electronically linked to the fine focus control.
The same holds true also for high-NA dry objectives because A (motor-driven) optical-correcting unit, such as the In-Focus
they are designed under the assumption that (unless embedded in a system (Infinity Photo-Optical Company, Boulder, CO), placed in
h = 1.52 medium) the specimen lies in an infinitely thin layer placed the parallel-beam region of a microscope can be used to change
directly against a coverslip whose thickness (generally 0.17 mm), the focal level and/or correct for residual spherical aberration
refractive index, and dispersion conform to specification. without displacing the objective lens. In fact we find that the unit
One approach to overcoming these problems is to switch to a can often improve the point-spread function, so that the z-axis dis-
water-immersion objective lens. Then the cumulative depth of the tribution of the 3D diffraction pattern becomes more symmetric,
water layer between the objective lens and the focused portion of even for “highly corrected” Plan Apochromatic objective lenses
the specimen should remain unchanged with focus. Whether the used following the manufacturer’s exact specifications (see
objective lens is designed for homogeneous water immersion or Chapters 7 and 9, this volume). Conversely, by appropriately
for use in the presence of a coverslip does not matter, so long as, linking the In-Focus drive with the objective (and condenser) lens
in the latter case, a coverslip with the proper specifications is used. motor drive(s), one can now to substantially improve the point-
In fact, however, even the small difference in h between physio- spread function even when focusing deeper, for example, into a
logical solutions, seawater, tissue, and pure water must be taken specimen residing in an aqueous medium.
into account. While this approach does overcome some of the aber-
ration problems, water-immersion lenses cannot be made with
NAs of much above 1.25 (because of the 1.33 refractive index of
Unintentional Beam Deviation
water). Several manufacturers now produces high-NA water- The intensity of each point in the final image from a confocal
immersion objectives with excellent correction, some with high microscope is designed to measure the amount of light trans-
transmissions for UV down to wavelengths of 340 nm. In our ex- mitted by the detector pinhole for the corresponding point in the
perience, the Nikon 60 ¥ 1.2 NA Plan Apochromatic, correction- specimen as it is being scanned. However, if the amount of light
collar-equipped water-immersion objective gave an impressive transmitted by the detector pinhole is modulated by factors not
DIC image of diatom frustule through a 220-mm-thick layer of related to the interaction of the illuminating point of light and the
specimen at that raster point, or if the confocality between the
entrance pinhole, the illuminated specimen point, and the detector
4
pinhole were to be transiently lost for any reason, one would obtain
Note also that the refractive index and dispersion of the immersion media
could also be significantly affected by temperature. Some high NA immer-
a false reading of the brightness at that point.
sion lenses are thus equipped with correction collars to compensate for these One such error could be introduced if a localized, lens- or
variations in addition to the thickness of the coverslip. prism-shaped region having a h different from that of the sur-
PHC1 28/11/2005 6:07 PM Page 16
roundings were present in the path of the scanning or imaging

beam. The scanning or imaging beam would then be refracted or 100
deviated and the intensity of light reaching the detector falsely
modified. Such a false signal could be difficult to distinguish from 80
variations in a genuine signal arising from a specimen point in the
focus plane. The plane of focus may also be distorted by the pres-
60
ence of such refracting regions (Pawley, 2002), so that one may
no longer be scanning a flat optical section through the specimen coherent
confocal
(Chapter 17, this volume). Moreover, as discussed in the previous 40
section, the diffraction pattern of each image point formed by an incoherent
oil-immersion lens can be distorted and displaced along the Z-axis 20 non-confocal
even when a layer of optically homogeneous aqueous medium is
present between the specimen and the coverslip [see Fig. 2–40
0
(after Gibson and Lanni, 1991) in Inoué and Spring, 1997]. 1 2 3 4 5 6 7 8
Attempts to correct some beam deviations by interferometric Normalized Spacing [l/(2 NA)]
measurements of the specimen have been published (Kam et al.,
2001). FIGURE 1.9. Experimental contrast transfer values measured as a function of
Clearly, vibration of the microscope, and even minor distor- the spatial period of line gratings using a laser beam-scanning microscope in
tions of the mechanical components that support the optics or the the confocal reflection mode (solid points) and in the non-confocal transmission
specimen, may introduce misalignment between the two pinholes mode (circles). In both imaging modes, PlanApo objective lenses (Nikon Inc.,
and what was supposed to be the confocal point in the specimen. Melville, NY) with numerical apertures (NAs) ranging from 0.45 to 1.4 and
This could lead to short-term periodic errors or longer-term drift. laser wavelengths (l) of 514.5 nm or 488 nm were used. Spatial periods are
Antivibration tables that isolate the instrument from building and expressed in units of the limiting wavelength, l/2(NA), to normalize the data
taken with different laser l and lenses of different NA. We call data presented
floor vibrations are commonly used to support confocal micro-
in this fashion the contrast transfer characteristic (CTC).
scopes. While such a support is useful and may be essential in Continuous lines are theoretical curves displaying calculated CTCs for the
some building locations, it does not eliminate the influence of air- coherent confocal and the incoherent non-confocal imaging modes. Compari-
borne vibration, which can in fact raise major havoc in microscopy son of the two CTC curves shows that, while the limiting resolutions are iden-
(G.W. Ellis, personal communication, 1966). Nor does it eliminate tical for both imaging modes, the contrast due to fine detail in the specimen is
the influence of thermal drift or vibration arising from the opera- maintained much better with confocal optics. (The microscope was a prototype
tion of the instrument. built by Hamamatsu Photonics Kabushiki Kaisha, Japan. It was used with the
Given the need to precisely maintain the confocal alignment detector pinhole diameter reduced to a small fraction of the Airy disk
and to use some form of mechano-optical scanning within the diameter. From Oldenbourg et al., 1993.)
instrument, a confocal microscope is especially susceptible to
vibration and problems of mechanical distortion. Indeed, once
We have confirmed these predictions by direct measurement
the complex optical, electro-optical, mechanical, and electronic
on test gratings that we fabricated by electron lithography with
systems have been appropriately designed, the success of one con-
spacings down to 0.1 mm. Indeed, with confocal optics equipped
focal instrument over another may well depend on its immunity
with a small exit pinhole, the contrast transfer efficiency rose to
to vibration, in addition to the friendliness of its user interface.
80% at twice (and reached 100% at three times) the cut-off
Yokogawa’s disk-scanning system, while with somewhat reduced
spacing. With incoherent non-confocal imaging, 80% contrast
confocal stringency, turns out to be remarkably immune to
transfer was not attained until four times and did not even reach
vibration problems.
the 100% transfer rate at eight times the cut-off spacing (Fig. 1.9)!
CONTRAST TRANSFER AND RESOLUTION IN SUMMARY

CONFOCAL VERSUS NON-CONFOCAL
MICROSCOPY • The limiting resolution of all microscopes depends on the l of
the light used and the NA of the objective and condenser
In addition to designing and successfully demonstrating the power lenses. Dirty or misaligned optics or vibration, or both, can
of stage-scanning confocal microscopy, Wilson and colleagues reduce the achieved resolution. Test resolution regularly, and
have extensively analyzed the theoretical foundations of confocal especially pay attention to the iris setting and full illumination
microscope imaging (see Wilson and Sheppard, 1984; Wilson, of the condenser aperture, to assure optimal performance.
1990; Chapters 11, 22, and 23 this volume). Their mathematical • A small detector pinhole in the confocal microscope is essen-
treatment leads to the somewhat surprising conclusion that the ulti- tial if the maximum optical sectioning capability and resolu-
mate limit of resolution (i.e., the cut-off spatial frequency or tion of the instrument are to be realized concurrently. Correct
spacing at which image contrast of periodic objects drops to zero) alignment and use of this control is very important. However,
obtainable with coherent confocal microscopy is identical with that a larger pinhole may be required to improve the signal when
for incoherent non-confocal microscopy. However, compared to there is limited light level, motion in the specimen, or fading
incoherent non-confocal optics, image contrast should rise much of fluorescence.
more sharply with coherent confocal optics as the spatial period is • By opening the confocal exit pinhole (to not much greater than
increased. Therefore, the practical resolution attained at threshold the Airy disk diameter), one loses the advantage of higher X–Y
contrast (i.e., the minimum contrast required for the spacing to be resolution in fluorescence microscopy but retains much of the
detected) was expected to be significantly greater with confocal capability for rejecting out-of-focus information while also
optics than with conventional non-confocal optics. gaining in the fluorescence signal.
PHC1 28/11/2005 6:07 PM Page 17
Åslund, N., Carlsson, K., Liljeborg, A., and Majlof, L., 1983, PHOIBOS, a
• The ultimate resolution limit of a CLSM in the reflection
microscope scanner designed for micro-fluorometric applications, using
mode is the same as that of a conventional light microscope, laser induced fluorescence. In: Proceedings of the Third Scandinavian
but the contrast that it produces from features is higher. Conference on Image Analysis, Studentliteratur, Lund, p. 338.
For fluorescence imaging, the resolution in a confocal micro- Åslund, N., Liljeborg, A., Forsgren, P.-O., and Wahlsten, S., 1987, Three
scope can be ~ 2 greater than with conventional microscopy, dimensional digital microscopy using the PHOIBOS scanner, Scanning
but only if the confocal detector pinhole is appreciably smaller 9:227–235.
than the Airy disk produced by a point fluorescent object. Baxes, G.A., 1984, Digital Image Processing: A Practical Primer, Prentice-
• CLSM is not a cure-all for all biological studies. Its sampling Hall, Englewood Cliffs, New Jersey.
Berek, M., 1927, Grundlagen der Tiefenwahrnehmung im Mikroskop, Marburg
speed is limited, and it does not lend itself to using either inter-
ference effects to produce contrast, such as phase or DIC, Sitzungs Ber. 62:189–223.
Born, M., and Wolf, E., 1980, Principles of Optics, 6th ed., Pergamon Press,
which have been found to be relatively innocuous to living
Oxford, England.
cells, or polarization contrast that can reveal fine structural Boyde, A., 1985a, Tandem scanning reflected light microscopy (TSRLM), Part
dynamics noninvasively. 2: Pre-MICRO 84 applications at UCL, Proc. Royal Microsc. Soc.
• Video microscopy can take advantage of the various types of 20:131–139.
interference and polarizing contrast not easily implemented in Boyde, A., 1985b, Stereoscopic images in confocal (tandem scanning)
the CLSM. Therefore, it is ideal for dynamic, high-resolution microscopy, Science 230:1270–1272.
observations of living specimens and for tracking the behav- Boyde, A., 1987, Colour-coded stereo images from the tandem scanning
ior of macromolecular assemblies. reflected light microscope (TSRLM), J. Microsc. 146:137–142.
• Holographic microscopy is a field with intriguing promise that Brakenhoff, G.J., Blom, P., and Barends, P., 1979, Confocal scanning light
microscopy with high aperture immersion lenses, J. Microsc. 117:219–
is so far beset by practical difficulties.
232.
• Two-photon microscopy and CLSMs provide high-confocal
Brakenhoff, G.J., van der Voort, H.T.M., van Spronsen, E.A., Linnemans,
stringency and, in general, are methods of choice for obtain- W.A.M., and Nanninga, N., 1985, Three dimensional chromatin distribu-
ing clear, high-resolution optical sections of the fluorescence tion in neuroblastoma nuclei shown by confocal scanning laser micro-
distribution in 3D fluorescent specimens. However, for scopy, Nature 317:748–749.
capturing well-resolved optical sections of highly dynamic Brakenhoff, G.J., van der Voort, H.T.M., van Spronsen, E.A., and Nanninga,
specimens, or for detection of sparsely distributed minute N., 1986, Three dimensional imaging by confocal scanning fluorescence
fluorescent objects, the newer disk-scanning confocal systems microscopy. In: Recent Advances in Electron and Light Optical Imaging
employing microlenses on the pinholes may well be the system in Biology and Medicine, Vol. 483 (A. Somlyo, ed.), Ann. N.Y. Acad. Sci.,
of choice. New York, pp. 405–414.
Brakenhoff, G.J., van Spronsen, E.A., van der Voort, H.T.M., and Nanninga,
N., 1989, Three dimensional confocal fluorescence microscopy, Methods
Cell Biol. 30:379–398.
ACKNOWLEDGEMENT Bright, G.R., Fisher, G.W., Rogowska, J., and Taylor, D.L., 1989, Fluorescence
ratio imaging microscopy, Methods Cell Biol. 30:157–192.
This chapter is an updated version of the chapter in the first and Cagnet M., Françon, M., and Thrierr, J.C., 1962, Atlas of Optical Phenomena,
second editions. We thank the staff of Yokogawa Electric Corpo- Springer-Verlag, Berlin.
ration for Figure 1.6 and Drs. James LaFountain and Rudolf Carlsson, K., Danielsson, P., Lenz, R., Liljeborg, A., Majlof, L., and Åslund,
Oldenbourg for making available the yet unpublished source mate- N., 1985, Three-dimensional microscopy using a confocal laser scanning
rial for Figure 1.8. We are most grateful to Professor Yoshinori microscope, Opt. Lett. 10:53–55.
Fujiyoshi of Kyoto University, whose support made the prepara- Castleman, K.R., 1979, Digital Image Processing, Prentice-Hall, Englewood
Cliffs, New Jersey.
tion of this chapter possible.
Castleman, K.R., 1987, Spatial and photometric resolution and calibration
requirements for cell image analysis instruments, Appl. Opt. 26:3338–
3342.
REFERENCES Castleman, K.R., 1993, Resolution and sampling requirements for digital image
processing, analysis, and display. In: Electronic Light Microscopy
Abbe, E., 1873, Beiträge zur Theorie des Mikroskops und der mikroskopischen (D. Shotton, ed.), Wiley-Liss Inc., New York, pp. 71–94.
Wahrnehmung, Schultzes Arc. f. Mikr. Anat. 9:413–468. Conchello, J.-A., Heym, J.P., Wei, J.L., and Lichtman, J.W., 1997, NOVEL
Abbe, E., 1884, Note on the proper definition of the amplifying power of a lens reflected light confocal profilometer, Proc. SPIE Conf. 2984:101–112.
or a lens-system, J. Royal Microsc. Soc. 4:348–351. Cox, G., and Sheppard, C., 1993, Effects of image deconvolution on optical
Agard, D.A., and Sedat, J.W., 1983, Three dimensional architecture of a poly- sectioning in conventional and confocal microscopes, Bioimaging
tene nucleus, Nature 302:676–681. 1:82–95.
Agard, D.A., Hiraoka, Y., Shaw, P., and Sedat, J.W., 1989, Fluorescence Cox, I.J., and Sheppard, C.J.R., 1983, Scanning optical microscope incorpo-
microscopy in three dimensions, Methods Cell Biol. 30: 353–377. rating a digital framestore and microcomputer, Appl. Opt. 22:1474–1478.
Allen, R.D., 1985, New observations on cell architecture and dynamics by Cox, I.J., and Sheppard, C.J.R., 1986, Information capacity and resolution in
video-enhanced contrast optical microscopy, Annu. Rev. Biophys. Bio- an optical system, J. Opt. Soc. Am. 3:1152–1158.
physical Chem. 14:265–290. Cremer, C., and Cremer, T. 1978, Considerations on a laser-scanning-
Allen, R.D., Travis, J.L., Allen, N.S., and Yilmaz, H., 1981a, Video-enhanced microscope with high resolution and depth of field, Microsc. Acta 81:31–44.
contrast polarization (AVEC-POL) microscopy: A new method applied to Davidovits, P., and Egger, M.D., 1971, Scanning laser microscope for biolog-
the detection of birefringence in the motile reticulopodial network of ical investigations, Appl. Opt. 10:1615–1619.
Allogromia laticollaris, Cell Motil. 1:275–289. Davidovits, P., and Egger, M.D., 1972, U.S. Patent #3,643,015, Scanning
Allen, R.D., Allen, N.S., and Travis, J.L., 1981b, Video-enhanced contrast, dif- Optical Microscope.
ferential interference contrast (AVEC-DIC) microscopy: A new method Denk, W., and Webb, W.W., 1987, Displacement fluctuation spectroscopy of
capable of analyzing microtubule-related motility in the reticulopodial the sensory hair bundles of mechanosensitive cells of the inner ear, Bull.
network of Allogromia laticollaris, Cell Motil. 1:291–302. Am. Phys. Soc. 32:645.
Amos, W.B., White, J.G., and Fordham, M., 1987, Use of confocal imaging in Denk, W., Strickler, J.H., and Webb, W.W., 1990, Two-photon laser scanning
the study of biological structures, Appl. Opt. 26:3239–3243. fluorescence microscopy, Science 248:73–76.
PHC1 28/11/2005 6:07 PM Page 18
Egger, M.D., 1989, The development of confocal microscopy, Trends Neurosci. scopy using a Nipkow scanner with microlenses — for 3-D imaging of
12:11. single fluorescent molecule in real-time, Bioimages 4:57–62.
Egger, M.D., and Petrán̆, M., 1967, New reflected-light microscope for viewing Ingelstam, E., 1956, Different forms of optical information and some interre-
unstained brain and ganglion cells, Science 157:305–307. lations between them. In: Problem in Contemporary Optics, Istituto
Ellis, G.W., 1966, Holomicrography: Transformation of image during recon- Nazionale di Ottica, Arcetri-Firenze, pp. 128–143.
struction a posteriori, Science 154:1195–1196. Inoué, S., 1981, Video image processing greatly enhances contrast, quality, and
Ellis, G.W., 1978, Advances in visualization of mitosis in vivo. In: Cell Repro- speed in polarization based microscopy, J. Cell Biol. 89:346–356.
duction, in Honor of Daniel Mazia (E. Dirksen, D. Prescott, and C.F. Fox, Inoué, S., 1986, Video Microscopy, Plenum Press, New York.
eds.), Academic Press, New York, pp. 465–476. Inoué, S., 1988, Progress in video microscopy, Cell Motil. Cytoskel. 10:13–17.
Ellis, G.W., 1979, A fiber-optic phase-randomizer for microscope illumination Inoué S., 1989, Imaging of unresolved objects, superresolution, and precision
by laser, J. Cell Biol. 83:303a. of distance measurement, with video microscopy, Methods Cell Biol.
Ellis, G.W., 1985, Microscope illuminator with fiber-optic source integrator, 30:85–112.
J. Cell Biol. 101:83a. Inoué S., 1994, Ultra-thin optical sectioning and dynamic volume investigation
Ellis, G.W., 1988, Scanned aperture light microscopy In: Proceedings of the with conventional light microscopy. In: Three-Dimensional Confocal
Forty-sixth Annual Meeting of EMSA, San Francisco Press, San Francisco, Microscopy (J.K. Stevens, L.R. Mills, and J. Trogadis, eds.), Academic
pp. 48–49. Press, San Diego, pp. 397–419.
Fay, F.S., Fogarty, K.E., and Coggins, J.M., 1985, Analysis of molecular dis- Inoué, S., and Inoué, T.D., 1986, Computer-aided stereoscopic video recon-
tribution in single cells using a digital imaging microscope. In: Optical struction and serial display from high-resolution light-microscope optical
Methods in Cell Physiology (P. De Weer and B.M. Salzberg, eds.), John sections. In: Recent Advances in Electron and Light Optical Imaging in
Wiley & Sons, New York. Biology and Medicine, Vol. 483 (A. Somlyo, ed.), Ann. N.Y. Acad. Sci.,
Flory, L.E., 1951, The television microscope, Cold Spring Harbor Symp. New York, pp. 392–404.
Quant. Biol. 16:505–509. Inoué, S., and Inoué, T.D., 2002, Direct-view high-speed confocal scanner —
Freed, J.J., and Engle, J.L. 1962, Development of the vibrating-mirror flying the CSU-10. In: Cell Biological Applications of Confocal Microscopy, 2nd
spot microscope for ultraviolet spectrophotometry, Ann. N.Y. Acad. Sci. ed. (B. Matsumoto, ed.), Academic Press, San Diego, pp. 87–123.
97:412–448. Inoué, S., and Oldenbourg, R., 1995, Optical instruments: Microscopes. In:
Fuchs, H., Pizer, S.M., Heinz, E.R., Bloomberg, S.H., Tsai, L-C., and Handbook of Optics, 2nd ed., Vol. 2 (M. Bass, ed.) McGraw-Hill, New
Strickland, D.C., 1982, Design and image editing with a space-filling 3D York, pp. 17.1–17.52.
display based on a standard raster graphics system, Proc. Soc. Photo. Opt. Inoué, S., and Spring, K., 1997, Video Microscopy, 2nd ed., Plenum Press, New
Instrum. Eng. 367:117–127. York.
Gibson, S.F., and Lanni, F., 1991. Experimental test of an analytical model of Inoué, S., Goda, M., and Knudson, R.A., 2001a, Centrifuge polarizing micro-
aberration in an oil-immersion objective len used in three-dimensional scope. II. Sample biological applications, J. Microsc. 201:357–367.
light microscopy, J. Opt. Soc. Am. A 8:1601–1613. Inoué, S., Knudson, R.A., Goda, M., Suzuki, K., Nagano, C., Okada, N.,
Goldstein, S.R., Hubin, T., Rosenthal, S., and Washburn, C., 1990, The VRIDI- Takahashi, H., Ichie, K., Iida, M., and Yamanaka, K., 2001b, Centrifuge
COM: a video rate image dissector confocal microscope. In: Optical polarizing microscope. I. Rationale, design, and instrument performance,
Microscopy for Biology (B. Herman and K. Jacobson, eds.), Wiley-Liss, J. Microsc. 201:341–356.
Inc., New York, pp. 59–72. Inoué, S., Shimomura, O., Goda, M., Shribak, M., and Tran, P.T., 2002, Fluo-
Gonzales, R.C., and Wintz, P., 1987, Digital Image Processing, 2nd ed., rescence polarization of green fluorescent protein (GFP), Proc. Natl. Acad.
Addison-Wesley, Reading, Massachusetts. Sci. 99:4272–4277.
Grego, S., Cantillana, V., and Salmon, E.D., 2001, Microtubule treadmilling in Kam, Z., Hanser, B., Gustafsson, M.G.L., Agard, D.A., and Sedat, J.W., 2001,
vitro investigated by fluorescence speckle and confocal microscopy, Computational adaptive optics for live three-dimensional biological
Biophys. J. 81:66–78. imaging, Proc. Natl. Acad. Sci. 98:3790–3795.
Gustafsson, M.G.L., 2000, Surpassing the lateral resolution limit by a factor Klar, T.A., Jacobs, S., Dyba, M., Egner, A., and Hell, S.W., 2000, Fluorescence
of two using structured illumination microscopy, J. Microsc. 198:82–87. microscopy with diffraction resolution barrier broken by stimulated emis-
Hamilton, D.K., and Wilson, T. 1984, Two-dimensional phase imaging in the sion, Proc. Natl. Acad. Sci. 97:8206–8210.
scanning optical microscope, Appl. Opt. 23:348–352. Koester, C.J., 1980, Scanning mirror microscope with optical sectioning char-
Hansen, E.W., 1986, Appendix II. In: Video Microscopy (S. Inoué, ed.), Plenum acteristics: Applications in ophthalmology, Appl. Opt. 19:1749–1757.
Press, New York, pp. 467–475. Kubota, H., and Inoué, S., 1959, Diffraction images in the polarizing micro-
Hard, R., Zeh, R., and Allen, R.D., 1977, Phase-randomized laser illumination scope, J. Opt. Soc. Am. 49:191–198.
for microscopy, J. Cell Sci. 23:335–343. Leith, E.N., and Upatnieks, J., 1963, Wavefront reconstruction with continu-
Harris, J.L., 1964, Diffraction and resolving power, J. Opt. Soc. Am. ous-tone objects, J. Opt. Soc. Am. 53:1377–1381.
54:931–936. Leith, E.N., and Upatnieks, J., 1964, Wavefront reconstruction with diffused
Hecht, E., 1987, Optics, 2nd ed., Addison-Wesley, Reading, Massachusetts. illumination and 3D objects, J. Opt. Soc. Am. 54:1295–1301.
Hell, S.W., and Wichmann, J., 1994, Breaking the diffraction resolution limit Lewin, R., 1985, New horizons for light microscopy, Science 230:1258–1262.
by stimulated emission: stimulated-emission-depletion fluorescence micro- Linfoot, E.H., and Wolf, E., 1953, Diffraction images in systems with an
scopy, Opt. Lett. 19:780–782. annular aperture, Proc. Phys. Soc. B 66:145–149.
Hell, S.W., Reiner, G., Cremer, C., and Stelzer, E.H.K., 1993, Aberrations in Linfoot, E.H., and Wolf, E., 1956, Phase distribution near focus in an aberra-
confocal fluorescence microscopy induced by mismatches in refractive tion-free diffraction image, Proc. Phys. Soc. B 69:823–832.
index, J. Microsc. 169:391–405. Maddox, P., Desai, A., Oegema, K., Mitchison, T.J., and Salmon, E.D., 2002,
Hellwarth, R., and Christensen, P., 1974, Nonlinear optical microscopic Poleward microtubule flux is a major component of spindle dynamics
examination of structure in polycrystalline ZnSe, Optics Comm. 12:318– and anaphase A in mitotic Drosophila embryos, Curr. Biol. 12:1670–
322. 1674.
Hoffman, R., and Gross, L., 1975, Modulation contrast microscopy, Appl. Opt. McCarthy, J.J., and Walker, J.S., 1988, Scanning confocal optical microscopy,
14:1169–1176. EMSA Bull. 18:75–79.
Hopkins, H.H., 1951, The concept of partial coherence in optics, Proc. Royal Minsky, M., 1957, U.S. Patent #3013467, Microscopy Apparatus.
Soc. Lond. 208A:263. Minsky, M., 1988, Memoir on inventing the confocal scanning microscope,
Hopkins, H.H., and Barham, P.M., 1950, The influence of the condenser on Scanning 10:128–138.
microscopic resolution, Proc. Royal Soc. Lond. 63B:737–744. Montgomery, P.O., Roberts, F., and Bonner, W., 1956, The flying-spot mono-
Ichihara, A., Tanaami, T., Isozaki, K., Sugiyama, Y., Kosugi, Y., Mikuriya, K., chromatic ultra-violet television microscope, Nature 177:1172.
Abe, M., and Uemura, I., 1996, High-speed confocal fluorescence micro- Nipkow, P., 1884, German Patent #30,105.
PHC1 28/11/2005 6:07 PM Page 19
Nomarski, G., 1955, Microinterférométre différentiel à ondes polarisées, Suzuki, T., and Hirokawa, Y., 1986, Development of a real-time scanning laser
J. Phys. Radium 16:S9–S13. microscope for biological use, Appl. Opt. 25:4115–4121.
Oldenbourg, R., 1996, A new view on polarization microscopy, Nature Tanasugarn, L., McNeil, P., Reynolds, G.T., and Taylor, D.L., 1984, Micro-
381:811–812. spectrofluorometry by digital image processing: Measurement of cyto-
Oldenbourg, R., and Mei, G., 1995, New polarized light microscope with pre- plasmic pH, J. Cell Biol. 89:717–724.
cision universal compensator, J. Microsc. 180:140–147. Tolardo di Francia, G., 1955, Resolving power and information, J. Opt. Soc.
Oldenbourg, R., Salmon, E.D., and Tran, P.T., 1998, Birefringence of single Am. 45:497–501.
and bundled microtubules, Biophys. J. 74:645–654. Tran, P.T., Marsh, L., Doye, V., Inoué, S., and Chang, F., 2001, A mechanism
Oldenbourg, R., Terada, H., Tiberio, R., and Inoué, S., 1993, Image sharpness for nuclear positioning in fission yeast based on microtubule pushing,
and contrast transfer in coherent confocal microscopy, J. Microsc. J. Cell Biol. 153:397–411.
172:31–39. Tsien, R.Y., 1989, Fluorescent indicators of ion concentration, Methods Cell
Pawley, J.B., 2002, Limitations on optical sectioning in live-cell confocal Biol. 30:127–156.
microscopy, Scanning 21:241–246. Volkmer, A., Cheng, J.-X., and Xie, X. S., 2001, Vibrational imaging with high
Petrán̆, M., Hadravsky, M., Egger, D., and Galambos, R., 1968, Tandem- sensitivity via epidetected coherent anti-Stokes Raman scattering
scanning reflected-light microscope, J. Opt. Soc. Am. 58:661–664. microscopy, Phys. Rev. Lett. 87:023901-1–023901-4.
Quate, C.F., 1980, Microwaves, acoustic and scanning microscopy. In: Scanned Waterman-Storer, C.M., and Salmon, E.D., 1997, Actomyosin-based retrograde
Image Microscopy (E.A. Ash, ed.), Academic Press, San Diego, pp. 23–55. flow of microtubules in the lamella of migrating epithelial cells influences
Sharnoff, M., Brehm, L., and Henry, R., 1986, Dynamic structures through microtubule dynamic instability and turnover and is associated with micro-
microdifferential holography, Biophys. J. 49:281–291. tubule breakage and treadmilling, J. Cell Biol. 139:417–434.
Sheppard, C.J.R., and Choudhury, A., 1977, Image formation in the scanning White, J.G., Amos, W.B., and Fordham, M., 1987, An evaluation of confocal
microscope, Optica 24:1051. versus conventional imaging of biological structures by fluorescence light
Sheppard, C.J.R., Gannaway, J.N., Walsh, D., and Wilson, T., 1978, Scanning microscopy, J. Cell Biol. 105:41–48.
Optical Microscope for the Inspection of Electronic Devices, Microcircuit Wijnaendts van Resandt, R.W., Marsman, H.J.B., Kaplan, R., Davoust, J.,
Engineering Conference, Cambridge. Stelzer, E.H.K., and Strickler, R., 1985, Optical fluorescence microscopy
Sher, L.D., and Barry, C.D., 1985, The use of an oscillating mirror for 3D dis- in three dimensions: Microtomoscopy, J. Microsc. 138:29–34.
plays. In: New Methodologies in Studies of Protein Configuration (T.T. Wilke, V., Gödecke, U., and Seidel, P., 1983, Laser-scan-mikroskop, Laser and
Wu, ed.), Van Nostrand-Reinhold, Princeton, New Jersey. Optoelecktron 15:93–101.
Shimizu, Y., and Takenaka, H., 1994, Microscope objective design. In: Wilson, T., 1985, Scanning optical microscopy, Scanning 7:79–87.
Advances in Optical and Electron Microscopy, Vol. 14 (C. Sheppard and Wilson, T., 1990, Confocal Microscopy, Academic Press, London.
T. Mulvey, eds.), Academic Press, San Diego, pp. 249–334. Wilson, T., and Sheppard, C., 1984, Theory and Practice of Scanning Optical
Shotten, D., ed., 1993, Electronic Light Microscopy: The Principles and Prac- Microscopy, Academic Press, London.
tice of Video-enhanced Contrast, Digital Intensified Fluorescence, and Wilson, T., Gannaway, J.N., and Johnson, P., 1980, A scanning optical micro-
Confocal Scanning Light Microscopy, John Wiley & Sons, New York. scope for the inspection of semiconductor materials and devices, J.
Shribak, M., Inoué, S., and Oldenbourg, R., 2002, Polarization aberrations Microsc. 118:390–314.
caused by differential transmission and phase shift in high-numerical- Xiao, G.Q., and Kino, G.S., 1987, A real-time confocal scanning optical micro-
aperture lenses: theory, measurement, and rectification, Opt. Eng. 41:943– scope. In: Proceedings of SPIE, Vol. 809, Scanning Imaging Technology
954. (T. Wilson and L. Balk, eds.), pp. 107–113.
Smith, L.W., and Osterberg, H., 1961, Diffraction images of circular self- Young, J.Z., and Roberts, F., 1951, A flying-spot microscope, Nature 167:
radiant disks, J. Opt. Soc. Am. 51:412–414. 231.
Squirrel, J.M., Wokosin, D.L., White, J.G., and Bavister, B.D., 1999, Long- Zernicke, V.F., 1935, Das Phasenkontrastverfahren bei der mikroskopischen
term two-photon fluorescence imaging of mammalian embryos without Beobachtung, Z. Tech. Phys. 16:454–457.
compromising viability, Nature Biotech. 17:763–767. Zworykin, V.K., 1934, The iconoscope — a modern version of the electric eye,
Stevens, J.K., Mills, L.R., and Trogadis, J., 1994, Three-Dimensional Confo- Proceedings of IRE 22:16–32.
cal Microscopy, Academic Press, San Diego. Zworykin, V.K., and Morton, G.A., 1954, Television: The Electronics of Image
Streibl, N., 1985, Three dimensional imaging by a microscope, J. Opt. Soc. Am. Transmission in Color and Monochrome, 2nd ed., John Wiley & Sons,
A 2:121–127. New York.

PHC1 - Foundations of Confocal Scanned Imaging in Light Micros

Uploaded by

Copyright:

Available Formats

PHC1 - Foundations of Confocal Scanned Imaging in Light Micros

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PHC1 - Foundations of Confocal Scanned Imaging in Light Micros

Uploaded by

Copyright:

Available Formats

PHC1 28/11/2005 6:07 PM Page 1

Foundations of Confocal Scanned Imaging in

Seldom has the introduction of a new instrument generated as 1.22l o

where rAiry is expressed as distance in the specimen plane.

Shinya Inoué • Marine Biological Laboratory, Woods Hole, Massachusetts 02543

TABLE 1.1. Historic Overviewa

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 3

TABLE 1.1. (Continued)

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 5

is projected onto the specimen by the “condenser.” The field of

Lewin, 1985). Boyde also provides striking stereoscopic images

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 7

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 9

FIGURE 1.4. Schematic of laser-scanning confocal micro-

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 11

reaching the specimen in order to avoid saturation of the fluo-

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 13

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 15

roundings were present in the path of the scanning or imaging

CONTRAST TRANSFER AND RESOLUTION IN SUMMARY

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 17

Foundations of Confocal Scanned Imaging in Light Microscopy • Chapter 1 19

You might also like