From MiToBo
Revision as of 11:02, 22 January 2019 by Moeller (talk | contribs) (→‎Usage)
Jump to navigationJump to search

Cell Boundary Extractor 2D

The Cell Boundary Extractor 2D is available since release version of MiToBo.

This operator aims to extract cell boundaries from microscope images with fluorescently-labeled cell membranes using vesselness enhancement filters.
For manual post-processing of sub-optimal segmentation outcomes you may take a look at our interactive label image editor.

Latest News

The CellBoundaryExtractor2D Plugin has been released in MiToBo and MiToBo-Plugins

Related Publications
  • B. Möller and K. Bürstenbinder,
    "Semi-automatic Cell Segmentation from Noisy Image Data for Quantification of Microtubule Organization on Single Cell Level".
    In Proc. of IEEE International Symposium on Biomedical Imaging (ISBI '19), Venice, Italy, April 2019, accepted.
Name of Plugin/Operator

(available since MiToBo version

Main features
  • automatic extraction of fluorescently-labeled cell membranes
  • gap closing based on Dijkstra shortest path techniques

To run the CellBoundaryExtractor2D perform the following steps:

  • install MiToBo by following the instructions on the Installation page

Important note:
the CellBoundaryExtractor2D operator relies for Dijkstra gap closing on version 1.2.0 of the external JGraphT library.
Fiji by default ships with an older version of this library which causes incompatibilites. To avoid breaking Fiji installations by activating MiToBo's update site we currently do not distribute the JGraphT library via our update site.
To use the CellBoundaryExtractor2D you need to manually install the jar:

  • Uninstall the old jar jgrapht-0.8.3.jar from your local installation via Help -> Update..., Advanced mode.
    Search for jgrapht and select as Status/Action the option Uninstall, then press Apply Changes and close Fiji.
    Note that this may render some other plugins like TrackMate to stop working, but there is currently no option to run them in parallel.
  • Download the new jar jgrapht-core-1.2.0.jar from here.
  • Copy the jar into the folder jars of your local Fiji installation and start Fiji again.

Afterwards the CellBoundaryExtractor2D can be invoked as follows:

  • run MiToBo and start the operator runner by selecting the menu item MiToBo Runner from Plugins -> MiToBo
  • in the selection menu navigate to 'de.unihalle.informatik.MiToBo.apps.cells2D' and select the operator CellBoundaryExtractor2D

This will bring up the operator window of the CellBoundaryExtractor2D.

  • Input data:
    To ease interplay with the CytoskeletonAnalyzer2D the operator expects a special organization of the input image data which is outlined in the figure below. All the data should be contained in a common top-level folder, here named "experiment".
    The images of different sub-groups (treatments/genotypes/proteins) should be found in separate sub-folders of this top-level folder, here named "group-1", "group-2" and so on. Note that the sub-folders named "results_features" can savely be ignored and don't have to be present for running the CellBoundaryExtractor2D. They will lateron be added by the CytoskeletonAnalyzer2D.

  • Output data:
    The operator generates for each image a label image with extracted cell regions as well as a file of cell boundaries in ImageJ ROI format. These files are stored in a new sub-folder termed "results_segmentation" in the sub-folder of each group. This way the output of the CellBoundaryExtractor2D can directly be specified as input for the CytoskeletonAnalyzer2D.
    Optionally, by checking the Option Show/save additional results? in the GUI additional images of intermediate results can be generated listed in the table below (<imageID> denotes the name of a given input image without ending).

Output File Name Description
<imageID>-label.tif Label image of extracted cell regions.
<imageID> ImageJ ROIs of extracted cell boundaries.
<imageID>-vessels.tif Visualization of the vesselness map.
<imageID>-vessels-binarized.tif Binarized vesselness map after applying Niblack thresholding.
<imageID>-vessels-binarized-filtered.tif Binarized vesselness map after filtering of, e.g., too small components.
<imageID>-skeleton-initial.tif Image of initial boundary skeleton after vesselness filtering and binarization.
<imageID>-skeleton-final.tif Image of final boundary skeleton after post-processing and gap closing.

  • Configuration Parameters:
Parameter Name Possible Values Description
Operation Mode BATCH Basic processing mode:
- BATCH: process all images in the sub-folders of the given top level folder
- SINGLE_IMAGE: process the currently active image only
Input Directory/Image Top-level folder containing sub-folders of all image groups to segment or input image name, respectively.
Cell Boundary Channel Channel with the fluorescently-labeled cell membranes.
Border Contrast
BRIGHT_ON_DARK Appearance of cell boundaries in images:
- BRIGHT_ON_DARK: boundaries are brighter than the background
- DARK_ON_BRIGHT: boundaries are darker than the background
Minimal Size of Cells Minimal size of cells, cells below this threshold are deleted.
Maximal Size of Cells Maximal size of cells, cells above this threshold are deleted.
Show/save additional results? Enables or disables display or storage, respectively, of additional intermediate results.


January 2019

  • Released first official version of CellBoundaryExtractor2D in MiToBo/MiToBo plugins