ZEISS Microscopy Knowledge Base
Help
ZEISS Microscopy Knowledge Base

Deconvolution

Airyscan Joint Deconvolution

The Airyscan Joint Deconvolution method provides an alternative processing of Airyscan data for even higher resolution. An unprocessed Airyscan file needs to be selected as an input file.

Airyscan Joint Deconvolution enables you to adjust the strength of the processing by defining the number of iterations or additionally the quality threshold. Note that overdoing the processing strength can result in artificially small structures, or splitting of solid structures.

Once processing is complete, review the output. Check the image Info view to see the number of iterations actually used, which can be informative especially if you have set a Quality Threshold. Save the processed image file with a new name to preserve the original unprocessed file for potential re-evaluation or different processing attempts. Ensure that the parameters are set according to the specifics of your sample and imaging requirements to avoid artifacts such as artificially small structures or splitting of solid structures.

Parameter

Description

Start with Last Result

Activated: Keeps the iteration result and adds additional iterations instead of starting the processing from the beginning again. Should be done with the same Pixel Upsampling setting.

Pixel Upsampling

Activated: Increases the sampling of output image by the factor of two for the best resolution.

Sample Structure

Sets the default value for typical sample types (different preset values for SR and MPLX data).

Sparse

Sets the default value for a sample with small structures and lots of black background (e.g. beads).

Standard

Sets the default value for a standard sample.

Dense

Sets the default value for a sample with lots of structures and little background.

Maximum Iterations

Sets the number of maximum iterations. This input is connected to the Quality Threshold parameter. If the Quality Threshold is set to 0, the number of iterations are processed as set here. If a threshold greater 0 is defined, the set number of iterations will not always be used, but processing stops if a certain resolution quality is achieved. The number of actual iterations is documented in the image Info view.

Quality Threshold

Sets the threshold for quality. This input is connected to the Maximum Iterations parameter.

Deblurring

Deblurring is a 2D background removal function based on the nearest neighbor algorithm. In the nearest neighbor method, the input is a 3D image where an in-focus slice is enhanced by subtracting its convolved neighbor(s). For Deblurring, the input is a 2D image. The neighbor is approximated using the 2D out-of-focus PSF from the microscope. It is then subtracted from the 2D input image the same way as the nearest neighbor algorithm.

Parameter

Description

Normalization

Only visible for Direct Processing.
Here you can select how the gray/color values that exceed or fall short of the value range should be dealt with. If you use this method with Direct Processing, only the Clip method is available and preselected.

Automatic

Automatically sets the gray levels that exceed or fall short of the predefined gray value range to the lowest or highest gray value (black or white). The effect corresponds to underexposure or overexposure. In certain circumstances some information may therefore be lost.

Clip

Normalizes the gray values automatically by clipping them to the available gray value range, for example values greater than the maximum grey value (e.g. 256 for 8 bit) are clipped to that value.

Strength

Sets the strength of the out of focus neighbor that is being subtracted. A higher value results in less background.

Sharpness

Sets the sharpness which is inversely related to the regularization strength. A higher value results in a sharper image. This parameter is only valid for the final deblurring that is done for further refinement of the subtracted image.

Blur Radius

Sets the PSF radius in pixels. The value of radius is proportional to the focal distance. A bigger radius results in less background.

Deconvolution (defaults)

Results with Direct Processing

Using deconvolution with Direct Processing may deliver different results than processing the image with the function on the Processing tab after acquisition. When processing the image offline with the Processing tab, the PSFs of all channels are stored in a single image document, with the maximum boundaries for all channels. When processing the image during acquisition with Direct Processing, each PSF is computed independently with its own boundaries. In Direct Processing, there is also no knowledge regarding which channel will come next or the maximum size of the PSF.

This method allows you to use four different algorithms for deconvolution, without any further settings. The following algorithms are provided in the Parameter tool:

Parameter

Description

Simple, very fast (Nearest Neighbor)

Only available, if a z-stack is selected as input image.
Executes the fast Nearest Neighbor method using default parameters.

Better, fast (Regularized Inverse Filter)

Executes the Regularized Inverse Filter algorithm for image enhancement.

Good, medium speed (Fast Iterative)

Executes the Fast Iterative restoration method.

Excellent, slow (Constrained Iterative)

Only available if you have licensed the Deconvolution functionality.
Executes the Constrained Iterative quantitative restoration method.

LSM DCV Processing

This method provides an alternative processing of confocal data for even higher resolution. An unprocessed confocal file needs to be selected as an input file.

LSM Deconvolution enables you to adjust the strength of the processing by defining the number of iterations or additionally the quality threshold. Note that overdoing the processing strength can result in artificially small structures, or splitting of solid structures.

Once processing is complete, review the output. Check the image Info view to see the number of iterations actually used, which can be informative especially if you have set a Quality Threshold. Save the processed image file with a new name to preserve the original unprocessed file for potential re-evaluation or different processing attempts. Ensure that the parameters are set according to the specifics of your sample and imaging requirements to avoid artifacts such as artificially small structures or splitting of solid structures.

Parameter

Description

Start with Last Result

Activated: Keeps the iteration result and adds additional iterations instead of starting the processing from the beginning again. Should be done with the same Pixel Upsampling setting.

Pixel Upsampling

Activated: Increases the sampling of output image by the factor of two for the best resolution.

Sample Structure

Sets the default value for typical sample types.

Sparse

Sets the default value for a sample with small structures and lots of black background (e.g. beads).

Standard

Sets the default value for a standard sample.

Dense

Sets the default value for a sample with lots of structures and little background.

Maximum Iterations

Sets the number of maximum iterations. This input is connected to the Quality Threshold parameter. If the Quality Threshold is set to 0, the number of iterations are processed as set here. If a threshold greater 0 is defined, the set number of iterations will not always be used, but processing stops if a certain resolution quality is achieved. The number of actual iterations is documented in the image Info view.

Quality Threshold

Sets the threshold for quality. This input is connected to the Maximum Iterations parameter.

PSF Wizard

Parameter

Description

Use Wizard

Activated: Opens the Measured PSF Wizard when applying the function to create a PSF.

Bad Pixel Correction

Employs a fully automatic detection and removal of spurious or hot pixels (also known as stuck pixels) in an image stack which might interfere with the PSF extraction procedure. It is based on the analysis of the gray level variance in the neighborhood of each pixel in the image. It is usually recommended to leave this parameter active.

Bead Proximity

Only visible if Use Wizard is deactivated.
Defines the distance between two neighboring beads in μm. A bead is excluded from the averaging if the distance to a neighboring bead is greater than the minimum distance set. A smaller value leads to the detection off fewer beads, albeit ones that are further apart, while a larger value leads to the detection of more beads, but with the risk that beads will partially overlap. Range between 2.1 and 20 µm.

Detection Sensitivity

Only visible if Use Wizard is deactivated.
Determines the sensitivity with which beads are being detected. Range is between -5 to 5. Smaller Values lower the sensitivity excluding beads with weaker staining and lower signal to noise ratio, higher values include also beads with weaker staining.

Average Beads

Only visible if Use Wizard is deactivated.
Activated: Produces an image of an individual, averaged bead.
Deactivated: Produces an image in which each found bead is centered, but saved in the R dimension. A slider for Rotation (R dimension) is displayed on the Dimensions tab.

Impressum
Carl-Zeiss-Strasse 22
73447 Oberkochen
Germany
Legal