This method allows you to use and individually configure two different algorithms for Apotome Plus. Two tabs are available for detailed configuration:
|
Parameter |
Description |
|
|---|---|---|
|
Algorithm |
Selects the Apotome deconvolution algorithm. |
|
|
– |
Fast Iterative (Joint) |
Uses an algorithm based on Deconvolution methods for structured illumination microscopy, with some enhancements as described in the technical note. It is faster and less memory intensive, and also purely using the image formation model. |
|
– |
Constraint Iterative (Joint) |
Uses an algorithm based on the Generalized approach for accelerated maximum likelihood based image restoration applied to three-dimensional fluorescence microscopy, but modified to allow for joint reconstruction and with enhancements as described in the technical note. It offers increased robustness to noise and mismatch between the theoretical and real PSF. Also, it offers more options (likelihood poisson and gaussian, and regularization), allowing choosing an algorithm optimized for specific image types, e.g. sparse and dense images. |
|
Enable Channel Selection |
Not possible in combination with Maximum Iterations and Quality Threshold. Activated: Applies the settings on a channel specific basis. This allows you to set parameters for each channel individually. A separate, colored tab for each of the channels is displayed. Deactivated: Applies the same settings to all channels of a multichannel image. |
|
|
2x Upsampling |
Activated: Allows you to extract additional information enabled by the SIM principle. This modality splits one pixel into four (2 vertical and 2 horizontal pixels), which allows the algorithm to work on a finer grid for deconvolution and reconstruct with higher resolution. A finer PSF is taken and processed in the same way as 1x resolution. Note that this will largely increase the computation times and requires a large CUDA-based GPU. |
|
|
Normalization |
Specifies how the data of the resulting image is handled if the gray/color levels exceed or fall short of the value range. |
|
|
– |
Clip |
Clips the values that exceed or fall short of the value range. Sets negative values to 0 (black). If the values exceed the maximum possible gray value of 65636 when the calculation is performed, they are limited to 65636 (pixel is 100% white). Results from different input images can be quantitatively compared with each other. |
|
– |
Automatic |
Normalizes the output image automatically. In this case the lowest value is 0 and the highest value is the maximum possible gray value in the image (gray value of 65636). The maximum available gray value range is always utilized fully in the resulting image. Results from different input images cannot directly be compared quantitatively with each other. |
|
– |
Factor |
Only visible if Clip is selected. |
|
Set Strength Manually |
Only available for Constrained Iterative and if for Regularization is at least Zero Order selected. Activated: Sets the desired degree of restoration with the slider. To achieve strong restoration and best contrast, move the slider towards Strong. To achieve lower restoration but smoother results, move the slider towards Weak. If the setting is too strong, image noise may be intensified and other artifacts, such as "ringing", may appear. Deactivated: Determines the restoration strength for optimum image quality automatically. This is recommended for widefield images and is therefore deactivated by default. The restoration strength is inversely proportional to the strength of so-called regularization. This is determined automatically with the help of Generalized Cross Validation (GCV). |
|
|
Convergence History |
Visible for Fast Iterative or Constrained Iterative algorithms. Displays the progress of the calculation as line graph. Several quality parameters are measured for each iteration and once either an optimum or the maximum allowed number of iterations is reached, the processing is stopped. This display allows you to observe directly how the iterative method affects the available data. It also shows how many iterations have been used and how much time is being used per iterations. |
|
|
Corrections |
To display parameters for image correction, click |
|
|
– |
Background |
Activated: Analyzes the background component in the image and removes it before the deconvolution calculation. This can prevent background noise being intensified during deconvolution. |
|
– |
Bad Pixel Correction |
Activated: 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 deconvolution result. |
|
– |
Fluorescent Decay |
Activated: Corrects bleaching of the sample during acquisition of the z-stack. |
|
– |
SIM Correction |
Activated: Removes stripe artefacts created by image acquisition and corrects for false phases in metadata. |
|
– |
Stronger Sectioning |
Activated: Applies a stronger optical sectioning to remove out-of-focus signals based on the multiplication of the optical sectioning data. |
To display advanced settings, click .
|
Parameter |
Description |
|
|---|---|---|
|
Likelihood |
Visible for Fast Iterative and Constrained Iterative algorithms. |
|
|
– |
Poisson (Richardson-Lucy) |
Only visible for the Fast Iterative algorithm. |
|
– |
Poisson |
Only visible for the Constrained Iterative algorithm. |
|
– |
Gauss |
Only visible for the Constrained Iterative algorithm. |
|
Regularization |
Only visible for the Constrained Iterative algorithm. |
|
|
– |
None |
No regularization is performed. |
|
– |
Zero Order |
Regularization based on G-difference, modeled on Tikhonov, but accelerated. |
|
– |
First Order |
Regularization based on Good's roughness. Under certain circumstances, more details are extracted from noisy data. It may be better suited to the processing of confocal data sets. |
|
– |
Second Order |
Regularization according to Tikhonov-Miller. Here higher frequencies are penalized more than in the case of Good's roughness. Results have a tendency to become overly smoothed. |
|
Optimization |
Visible for Fast Iterative and Constrained Iterative algorithms. |
|
|
– |
Analytical (Newton Raphson) |
Only visible for the Constrained Iterative algorithm. |
|
– |
Line Search |
Only visible for the Constrained Iterative algorithm. |
|
– |
Numerical Gradient |
Only visible for the Fast Iterative algorithm. |
|
First Estimate |
Visible for Fast Iterative and Constrained Iterative algorithms. |
|
|
– |
Input Image |
The input image is used as the first estimate of the target structure (default). |
|
– |
Last Result Image |
The result of the last calculation is used to estimate the next calculation. This can speed up a calculation that is repeated using slightly different parameters. |
|
– |
Mean of Input |
No estimate is made, the mean gray level of the input image is being used. This is the most rigid application of deconvolution. It should be chosen for confocal images, where the data sampling can be quite sparse. The computation time will increase, but missing information can be recovered from the PSF. |
|
– |
Zero Values |
Only visible for the Constrained Iterative algorithm. |
|
Maximum Iterations |
Visible for Fast Iterative and Constrained Iterative algorithms. |
|
|
Quality Threshold |
Only visible for the Fast Iterative and Constrained Iterative algorithms. |
|
|
Since Apotome Plus only supports GPU, the following two options cannot be edited: |
||
|
GPU Acceleration |
Only visible if a suitable (NVIDIA, CUDA based) graphics card is installed in your PC. The checkbox is then activated by default. Activated: Uses GPU processing. |
|
|
GPU Tiling |
Only available for very large images that exceed the available graphic card memory. Activated: With this function the image is split up in smaller portions which fit into the memory of the graphic card. The function automatically determines into how many tiles the image must be split to allow maximum usage of the graphics card. The resulting tiles are automatically stitched together for the final output result. Deactivated: No tiling is performed, however, in this case only certain sub-functions of deconvolution can run on the graphics card and the speed increase compared to CPU processing will be lower. The image quality might be higher than with tiling because there is no need for stitching. |
|
All key parameters for generating a theoretically calculated Point Spread Function (PSF) are displayed on this tab.
Usually, images (with file type *.czi) that have been acquired with ZEN automatically contain all microscope parameters, meaning that you do not have to configure any settings on this tab. Therefore, most parameters are grayed out in the display. It is possible, however, that as a result of an incorrect microscope configuration values may not be present or may be incorrect. You can change them here. The correction of spherical aberration can also be set here.
|
Parameter |
Description |
|---|---|
|
Get PSF Parameters from Input Image |
Activated: The parameters of the PSF are taken directly from the input image and the respective parameters are displayed and grayed out (not editable). This is the default setting. Deactivated: Enables you to manually change the PSF parameters. |
The most important microscope parameters for PSF generation that are not channel-specific are displayed in this section.
If you enter incorrect values, this can lead to incorrect calculations. If the values here are obviously wrong or values are missing, check the configuration of your microscope system.
|
Parameter |
Description |
|---|---|
|
NA Objective |
Displays the numerical aperture of the objective. |
|
Immersion |
Displays the refractive index of the immersion medium. Note that this can never be smaller than the numerical aperture of the objective. You can make a selection from typical immersion media in the dropdown list next to the input field. |
|
Scale Lateral |
Displays the geometric scaling in the X/Y direction. |
|
Scale Axial |
Displays the geometric scaling in the Z direction. |
Only visible if the Show All mode is activated.
|
Parameter |
Description |
|
|---|---|---|
|
Phase Ring |
If you have acquired a fluorescence image using a phase contrast objective, the phase ring present in the objective is entered here. This setting has significant effects on the theoretical Point Spread Function (PSF). |
|
|
PSF Generation |
Selects the model for calculating the PSF. |
|
|
– |
Scalar Theory |
The wave vectors of the light are interpreted as electrical field = intensity and simply added. This method is fast and is sufficient in most cases (default setting). |
|
– |
Vectorial Theory |
The wave vectors are added geometrically. However, the calculation takes considerably longer. |
|
Z-Stack |
This field can only be changed if it was not possible to define this parameter during acquisition, e.g. because the microscope type was unknown. It describes the direction in which the z-stack was acquired. Note that this setting is only relevant, if you are using the spherical aberration correction. |
|
|
– |
Descending |
The z-stack descends away from the objective. |
|
– |
Ascending |
The z-stack ascends towards the objective. |
Only visible if the Show All mode is activated.
Here you can select whether you want spherical aberration to be taken into account and corrected during the calculation of the PSF. As with the other PSF parameters, most values are extracted automatically from the information about the microscope that is saved with the image during acquisition. The input option is therefore inactive.
|
Parameter |
Description |
|---|---|
|
Enable Correction |
Activated: Uses the correction function. All options are active and can be edited. |
|
Embedding Medium |
Select the used embedding medium. |
|
Refractive Index |
Displays the refractive index of the selected embedding medium. Enter the appropriate refractive index if you are using a different embedding medium. |
|
Manufacturer |
Displays the manufacturer, if known. |
|
Depth Variance |
When Aberration correction is activated, it is also possible to enable the creation of depth variant PSFs. This method allows for dramatic improvements in image restoration of thicker samples by creating axially variant theoretical PSFs as a function of the distance to the coverslip and the refractive index of the mounting medium. Activated: Uses depth variant aberration correction. In the spin box edit field you can define how many PSFs should be generated. The more PSFs you create, the better the results, but selecting many PSFs increases the processing time. You should choose at least 3 PSFs. In the dropdown list you can choose between the PCA method (Primary Component Analysis, M. Arigovindan et al., 2005, IEEE Transactions on Image Processing 14. nr. 4 p.450ff) which is best suited for constrained iterative and fast iterative method and the Strata method (Myneni and Preza, Frontiers in Optics 2009, Optical Society of America, paper CThC4.), which is best for regularized inverse filter and Richardson Lucy iterative deconvolution. |
|
Distance to Cover Slip |
Displays the distance of the acquired structure from the side of the cover slip facing the embedding medium. Half the height of the z-stack is assumed as the initial value for the distance from the cover slip. The value can be corrected if this distance is known. If possible, this distance should be measured. Note: Use Ortho View and the Distance Measurement option to define the distance of the sample to the coverslip. It is also important to estimate the position of the glass/embedding medium interface as precise as possible. If the z-stack extends into the coverslip, the determined range of the stack which reaches into the glass should be entered as a negative value. Example: Z-stack is 26 µm thick, glass/medium interface is positioned at 9 µm distance from the first plane of the stack. Resulting value for Distance to cover slip: - 9.0 µm. |
|
Cover Slip Type |
Commercially available cover slips are divided into different groups depending on their thickness (0, 1, 1.5 and 2), which you can select from the dropdown list. Cover slips of the 1.5 type have an average thickness of 170 µm. In some cases, however, the actual values can vary greatly depending on the manufacturer. For best results the use of cover slips with a guaranteed thickness of 170 µm is recommended. Values that deviate from this can be entered directly in the input field. |
|
Cover Slip Ref. Index |
Selects the material that the cover slip is made of. The corresponding refractive index is displayed in the input field next to it. |
|
Working Distance |
Displays the working distance of the objective (i.e. the distance between the front lens and the side of the cover slip facing the objective). The working distance of the objective is determined automatically from the objective information, provided that the objective was selected correctly in the MTB 2011 Configuration program. You can, however, also enter the value manually. |
|
Override |
Only active if Enable Correction is activated. |
In this section you find all settings that are channel specific. This means that they can be configured differently for each channel.
|
Parameter |
Description |
|---|---|
|
Illumination |
Displays the excitation wavelength for the channel dye [in nm] by using the peak value of the emission spectrum. The color field corresponds to the wavelength (as far as possible). |
|
Detection |
Displays the peak value of the emission wavelength for the channel dye. The color corresponds to the wavelength (as far as possible). |
|
Sampling Lateral |
Depends on the geometric pixel scaling in the X/Y direction and displays the extent of the oversampling according to the Nyquist criterion. The value should be close to 2 or greater in order to achieve good results during DCV. As, in the case of widefield microscopes, this value is generally determined by the objective, the camera adapter used and the camera itself, it can only be influenced by the use of an Optovar. With confocal systems, the zoom can be set to match this criterion. |
|
Sampling Axial |
Depends on the geometric pixel scaling in the Z direction and displays the extent of the oversampling according to the Nyquist criterion. The value should be at least 2 or greater in order to achieve good results during DCV. This value is determined by the increment of the focus drive during acquisition of Z-stacks and can therefore be changed easily. |
Displays advanced microscope information that influences the form of the PSF in a channel-dependent way:
|
Parameter |
Description |
|---|---|
|
Illumination |
Selects the illumination method with which the data set has been acquired. In the event that a Conventional Microscope has been entered under the microscope parameters, the following options are available here: Epifluorescence, Multiphoton Excitation and Transmitted Light. In the case of confocal microscopes, Epifluorescence is the only option. |
|
Image Formation |
Displays whether the imaging was incoherent (Conventional Microscope) or coherent (Laser Scanning Microscope). |
|
Lateral Resolution |
Displays the lateral resolution of the calculated PSF. |
|
Axial FWHM |
Displays the FWHM (Full Width Half Maximum) as a measure of the axial resolution of the PSF. |
This section shows you the PSF that is calculated for a channel based on the current settings. If you select the Auto Update checkbox, all changes made to the PSF parameters are applied immediately to the PSF view. This makes it possible to check quickly whether the settings made meet your expectations. You can extract the PSF from the image via right-click menu (PSF Snapshot) which opens the resulting new PSF document in the center screen area.