Image processing pipelines in Image Analyst MKII
Image Analyst MKIIPipelines Glossary
HomeWorkflowImage Processing BasicsFunctions GlossaryPipelines GlossaryProtocolsQuick How ToSearch

Measure of intensities in single-cells in two channels with spectral unmixing

Parameters:
Name # Type Description
Channel Number #1 1 integer Channel number associated with the image, as appears in the Multi-Dimensional Open Dialog. Use Edit/Rename to view or change channel number.
Channel Number #2 2 integer Channel number associated with the image, as appears in the Multi-Dimensional Open Dialog. Use Edit/Rename to view or change channel number.
Local background subtraction by median rolling ball 3 boolean Rolling ball-style background subtraction will be performed: the median of the image is subtracted from the image.
Rolling ball: Median filter size (pixels) 4 integer The median is calculated over a width x width rectangular area around each pixel of the image.
Background level (percentile) 5 real Background is calculated as frame-by-frame median of those pixels that are the darker than this percentile of image histogram for the duration of the entire recording. This has no effect if rolling ball is used.
Spectral Unmix Coefficient Matrix 6 string To determine spectral crossbleed coefficients use the Tools/Calculate Crossbleed Correction Factor main menu point, or the Math/Blind Spectral Unmix with NMF function. See more on the layout of the Spectral Unmix Coefficient Matrix in the description of Math/Spectral Unmix. The default ={{1,0},{0,1}} means that there is no crossbleed.
Image stabilizer compares to first frame 7 boolean “Yes” is the fastest way of image stabilization. Turn it off if the image changes a lot during the time lapse, and it is not possible to compare later frames to the first frame any more.
Image stabilizer period (frames) 8 integer 1 if comparing to the first frame. Use more frames if not comparing to the first frame and the view field drifts slowly.
ROIs: Nucleus diameter (pixels) 9 integer Diameter of the nucleus in pixels. Nuclei ranging around this size will be selected.
ROIs: Cell diameter (pixels) 10 integer Diameter of the cell in pixels in the combined TMRM+FLIPR projection image. Cells ranging around this size will be selected.
ROIs: Margin around cells (pixels) 11 integer Cell boundaries will be extended by this amount.
ROIs: Debris cutoff (percentile) 12 real This percentile of the image histogram sets the intensity value where the maximum of the Look Up Table (LUT) is scaled. Use -1 to override this with fixed value set below at "Max value".
ROIs: Minimum cell fluorescence (%) 13 real Cells dimmer than this in filtered rescaled TMRM+FLIPR projection images will be rejected. Increase this value if debris dimmer than the cells is detected.
ROIs: Sensitivity of cell boundaries (% of max fluorescence) 14 real For "Bound locally", the boundary of each object is determined at this % of the maximal intensity of the object relative to its neighborhood. For "Bound uniformly" this is a pixel intensity value.
ROIs: Weld segments into round objects 15 boolean Weld touching segments if they form a rounder object together. Use this to avoid objects fragmenting into multiple segments.
Plot normalization 16 string
Show results in microplate format 17 boolean Organizes data in table according to microplate well associations in additional worksheets.
Description:
This pipeline performs all image processing tasks required for calibration of a TMRM / FLIPR (tetramethylrhodamine methyl ester and Plasma Membrane Potential Indicator; Molecular Devices FLIPR Plasma membrane potential kit) fluorescence time lapse recording to millivolts. In addition, post-hoc immunocytochemistry (ICC) images (showing the same view field after immunofluorescence staining) are used to selectively circle a specific cell type. Processed images and calculated ROIs are passed to the Membrane Potential Calibration Wizard. Using pre-defined configuration, this pipeline is capable of unsupervised processing of multi-stage position mitochondrial membrane potential assays.
The pipeline requires two Multi-Dimensional Open dialogs as input. One containing the live-cell time lapse of TMRM and FLIPR fluorescence, tagged as #1 in the top right corner of the dialog. The second dialog (tagged as #2) opens the post-hoc immunocytochemistry, containing a nuclear stain channel and a qualifier stain (e.g. insulin for beta-cells).
The pipeline corrects for misalignment between channels on the frame-by-frame basis, subtracts background, performs spectral unmixing with pre-defined coefficients, and stabilizes the time lapse (registers frames in time). To eliminate background the “Median of pixels below percentile of max projection” with the percentile given at “Background level” is used.
To determine spectral crossbleed coefficients use the Tools/Calculate Crossbleed Correction Factor main menu point, or the Math/Blind Spectral Unmix with NMF function. See more on the layout of the Spectral Unmix Coefficient Matrix in the description of Math/Spectral Unmix.
Then ROIs are generated automatically using the nuclear stain as seeds, following the shape of the maximum intensity projection of the summed FLIPR+TMRM image. Therefore if a cell moves during the experiment, the ROI will contain the cell for the whole duration of the recording. The calibration is not affected by the amount of background recorded in the ROI.
The immunofluorescence stain is used the classify ROIs, removing ones with lower than a threshold fluorescence.
Prepared TMRM and FLIPR images with immunofluorescence positive ROIs are passed to the Membrane Potential Calibration Wizard. Calibration can be performed by manual configuration of the wizard, or using a previously saved configuration file to provide automation.
The Membrane Potential Calibration Wizard implements the calibration technique published by Gerencser et al. in J Physiol. 2012 590:12 2845-71.

Version history:
V2
Segment welding has been enabled and separation of segments has been disabled.
No clipping during intensity scaling for seed calculation.
Intensity classifiers are applied to the seeds, as well as to the secondary segmentation.
Improved background subtraction
V3
Function parameter rename: Minimum cell ICC fluorescence (%),
Added ROIs: Sensitivity of cell boundaries (% of max fluorescence) parameter