In a conventional epifluorescence microscope, short wavelength light (e.g. blue light when fluorescein being used as the fluorophore) is reflected by a chromatic reflector through the objective and bathes the whole of the specimen in fairly uniform illumination. The chromatic reflector has the property of reflecting short wavelength light and transmitting longer wavelength light. Emitted fluorescent light (e.g. longer wavelength, green light in the case of fluorescein) from the specimen and passes straight through the chromatic reflector to the eyepiece.
In a confocal imaging system a single point of excitation light (or sometimes a group of points or a slit) is scanned across the specimen (see animation 1). The point is a diffraction limited spot on the specimen and is produced either by imaging an illuminated aperture situated in a conjugate focal plane to the specimen or, more usually, by focusing a parallel laser beam. With only a single point illuminated, the illumination intensity rapidly falls off above and below the plane of focus as the beam converges and diverges, thus reducing excitation of fluorescence for interfering objects situated out of the focal plane being examined. Fluorescent light (i.e. signal) passes back through the dichroic reflector and then passes through a pinhole aperture situated in a conjugate focal plane to the specimen. Any light emanating from regions away from the vicinity of the illuminated point will be blocked by the aperture, thus providing yet further attenuation of out-of focus interference (see animation 2, and animation 3). Light passing through the image pinhole is detected by a photodetector. Usually a computer is used to control the sequential scanning of the sample and to assemble the image for display onto a video monitor. Most confocal microscopes are implemented as imaging systems that couple to a conventional microscope.
In summary, a confocal imaging system achieves out-of-focus rejection by two strategies: a) by illuminating a single point of the specimen at any one time with a focussed beam, so that illumination intensity drops off rapidly above and below the plane of focus and b) by the use of blocking a pinhole aperture in a conjugate focal plane to the specimen so that light emitted away from the point in the specimen being illuminated is blocked from reaching the detector. Confocal imaging can offer another advantage in favorable situations (small pinhole size, bright specimen): the resolution that is obtained can be better by a factor of up to 1.4 than the resolution obtained with the microscope operated conventionally.
A confocal microscope can be used in reflection mode and still exhibit the same out-of-focus rejection performance. There are some biological applications of this mode of operation such as for detecting gold particles used in immuno-gold labeling. However, in the vast majority of biomedical applications they are used for fluorescence imaging.
Applications of Confocal Microscopy
In general, a confocal microscope that is set up correctly will always give a better image than can be obtained with a standard epifluorescence microscope. All this improvement essentially comes from the rejection of out-of-focus interference. The improvement can vary between marginal, in the case of very flat specimens like chromosome squashes, to spectacular, in the case of large, whole-mount specimens such as embryos. Indeed in the latter case it is often impossible to distinguish any interior detail with conventional microscopy yet obtain a perfectly clear image of an optical section using confocal imaging (example picture). The technique is therefore applicable to all fluorescence microscopy applications.
The current (1999) generation of confocal microscopes generally have laser excitation and the capability to simultaneously detect two or more emitted colors. The simultaneous detection of two or more colors has the advantage that registration problems associated with sequentially imaging are avoided and the dose of irradiation given to the specimen is reduced. Popular excitation sources are the argon ion laser that emits at 488nm and 514nm and the argon/krypton mixed gas laser that gives three useful spectral lines for excitation at 488nm, 568nm and 647nm. These wavelengths cover many of the commonly used fluorophores with the notable exception of the popular DNA stains DAPI and Hoescht 33258 which require UV excitation. Large-frame, high-power argon ion lasers can provide UV lines but are rather expensive. Also, the axial chromatic aberration present in many objectives between the UV and visible wavelengths often gives rise to disappointing results.
Most confocal imaging systems provide adjustable pinhole blocking apertures. This enables a tradeoff to be made in vertical resolution and sensitivity. A small pinhole gives the highest resolution and lowest signal and vice versa.
The photodetectors most commonly used are photomultipliers. These have reasonable sensitivity in the blue regions of the spectrum but markedly fall of in sensitivity in the red. Charge coupled device (CCD) detectors offer higher quantum efficiency (sensitivity) and extended red responses, but are not compatible with the high readout rates generated by the current generation of point-scanning confocal imaging systems. Avalanche photodiodes offer considerable promise as detectors for confocal imaging as they are fast and have quantum efficiencies approaching that of CCDs.
The image in a beam-scanning confocal microscope is assembled in digital form in a frame store or in computer memory. It is therefore amenable to digital image processing techniques. Most manufacturers provide a set of basic image processing functions, such as contrast and black level manipulation, pseudocolouring and spatial filtering (to provide edge enhancement or smoothing to an image). Images may be conveniently archived onto hard disk or replaceable optical disks. Hard copy may be obtained by photographing the display screen or be means of a video printer.
Limitations of point-scanning confocal microscopy
Point-scanning microscopes, when used with high numerical aperture lenses, have an inherent speed limitation in fluorescence. This arises because of a limitation in the amount of light that can be obtained from the small volume of fluorophore contained within the focus of the scanned beam (less than a cubic micron). At moderate levels of excitation, the amount of light emitted will be proportional to the intensity of the incident excitation. However, fluorophore excited states have significant lifetimes (in the order if a few nanosecond). Therefore, as the level of excitation is increased, the situation eventually arises when most of the fluorophore molecules are pumped up to their excited state and the ground state becomes depleted. At this stage the fluorophore is saturated and no more signal may be obtained from it by increasing the flux of the excitation source. Most commercial scanning beam confocal microscopes have laser excitation sources that give around 10 mw in the major spectral lines. When the spectral line is near the excitation peak of the fluorophore being used (e.g. the 488nm argon line and fluorescein) and a high numerical aperture lens is used (>1.0 NA) this power level will cause saturation giving image degradation. Better images will be obtained by reducing the power by a factor of 10 or 100. This limits the speed which an image with a given signal-to noise ratio can be acquired. Typically, about 5 -10 seconds of integration is required with an average immunofluorescence preparation.
Parallel beam confocal imaging systems
The speed limitations imposed by fluorophore saturation can be overcome by parallelism. An array of beams can be used in parallel, or a line of illumination can be used which can be thought of as a one dimensional array of points. Multiple-beam confocal microscopes have a higher speed potential than point-scanners because their inherent parallelism avoids fluorophore saturation enabling higher levels of excitation to be used. Most multiple beam confocal configurations optically reconstruct the image. This allows the use of high sensitivity CCD detectors giving extended red response. Multiple-beam confocal systems are likely to find applications in areas such as the imaging of high-speed intracellular events such as the visualization of calcium ion transients.