Resolution in the x, y plane

As in wide-field microscopy, resolution at the focal plane is determined by the diameter of the Airy disc, and therefore is said to be diffraction limited. Instrument parameters such as scan rate and pin-hole diameter can be set to achieve maximum resolution. If the sample has low light emission the pinhole can be opened greater than one Airy diameter to allow more light to the detector. In this case, however, resolution will decrease (slightly) at the expense of being able to image the sample at all.

Within the resolution of the CLSM (confocal laser scanning microscope) set parameters, any variation in actual sample intensities will be averaged, or integrated, into one intensity value in any image voxel. It is important, therefore, to scan a sample at a resolution that meets the criteria of your experiment and conforms to the anatomy of the sample. Small sample structures can be missed if the step size is too high or the voxel (3D pixel) too large. The operator must consider the size of the objects of interest and set pixel size and step size to appropriate values in order to collect enough information to faithfully reconstruct the object in question. In doing so, one fulfills the Nyquist theorem of sampling.

The sampling frequency (f) required to reconstruct an object of spatial frequency (d), or phenomenon of temporal frequency, must be at least d/2 in order to distinguish object information from background. In an image, small objects are high-frequency and large objects are low-frequency. Thus, to discern an object that is, say, 0.5 µm (f-0.5µm) the acquisition pixel size must be set to at most 0.25 µm. "Oversampling" (i.e., a smaller pixel size, say 0.1µm) would capture the same information, but take a longer time and result in a larger digital image file.

Resolution in a confocal system

The spreading of light from a luminous object: The Point Spread Function

Light emanating from a point source is focused by a lens system, not to an identical point in the image plane, but to a bright spot, called the Airy disc, surrounded by a series of concentric bright and dark rings in the x and y plane, and to complex, elongated cones of light in the axial, or z-dimension, above and below the plane of focus (right). The axial spreading of light from an illuminated point-source is called the point spread function (PSF). The confocal pinhole acts to reduce the effect of diffraction on image formation. Eliminating outer rings increases overall resolution. However, there is a tradeoff between decreased pinhole size (increasing resolution) and increased detector noise (because of the greatly decreased amount of light passing through the very small pinhole).

PSF of a self-luminous object

Axial resolution

The depth of a voxel (along the z-dimension or optical axis) has typically only half the resolution of the pixel in the x- and y-dimension. Factors affecting axial resolution are the objective numerical aperture (NA) and pinhole diameter. Increasing the NA and/or decreasing the diameter of the pinhole will increase the z‑resolution. In any case, the elucidation of sample depth information is always less than the x- and y-resolution due to the blurring effect of the PSF. In practice, the maximum resolution in Z (axial) that can be realized in a confocal microscope system is about 0.8µm; 2–3x worse than in the xy-dimension.

Overall sample resolution

Another factor that can contribute to decreased sample resolution is the optical sectioning rate. This rate is a measure of the spacing between successive optical sections (step-size) and not the actual “thickness” of the optical section itself. When taken at the same resolution as z, an optical section series (contiguous optical sections through z)  will contain all possible sample information. However, this rate of data collection may not be necessary to discern a given biological phenomenon. Decreasing the number of optical sections through a given axial dimension may yield similar results, and will decrease the total scanning time and the resulting computer disk space required to store the images.

Confocal lateral resolution
WF lateral resolution
Where f is the sampling frequency

d is the size of the smallest object to be imaged