Corsen ImageJ Plugin

The Corsen analysis is a two steps process:

  • 3D segmentation
  • Distance computation

Step 1 is performed using the image processing software ImageJ. We provide a plug-in for ImageJ (the Corsen plug-in) that allow converting an image stack to a ".par" file that resumes all the informations about the 3D segmented particles (areas, volumes, coordinates...).

As you need two particles types to compute distances, you must perform 3D segmentations for both of them using Corsen plug-in.

Installation

If the ImageJ software is not available on your computer, you can freely download it (It's a domain public program) on the ImageJ website.

To install Corsen ImageJ plug-in you only need to put the corsen-plugin.jar archive in the plugins folder of your ImageJ installation (usualy the default ImageJ directory installation under Microsoft Windows is located in c:\Program Files\ImageJ . See the download section for more information.

Allocating more memory to ImageJ

When you work with heavy images in ImageJ, the execution of the Corsen plug-in will require more memory. To allow the Java Virtual Machine that run ImageJ to use more memory, you need to edit the following files :

  • For Linux/Unix : run script shell in the ImageJ installation directory. Set the -Xmx parameter the define the maximal amount of memory. The following sample will set maximal memory use by ImageJ to 512 MB:
  • #!/bin/sh
    ./jre/bin/java -Xmx512m -jar ij.jar
    
  • For Windows : imagej.cfg in ImageJ installation directory (usualy C:\Program Files\ImageJ ). The third line contains arguments to the JVM to launch ImageJ. Set the -Xmx parameter the define the maximal amount of memory. Note that with Windows 32 bits versions, the maximum size of memory allocation is 1610 MB. The following sample will set maximal memory use by ImageJ to 1610 MB:
  • C:\PROGRA~1\ImageJ\
    C:\Program Files\Java\jre1.6.0_05\bin\javaw.exe
    -Xmx1610m -cp ij.jar;C:\PROGRA~1\ImageJ\ij.jar ij.ImageJ
    

For Mac OS X users, see the Image J documentation for more informations.

Usage

Since Corsen segmentation is based on a simple threshold, we advise Corsen users to preprocess images (e.g. signal enhancement, removing of background) before using Corsen. The quality of the Corsen 3D segmentation will be better (see preprocessing section).

Corsen ImageJ plugin can only handle the following images types :

  • Stacks of 8-bit grayscale images.
  • Stacks of 16-bit images.

Before running Corsen ImageJ plug-in you must define a threshold using the Theshold tool of ImageJ (menu Image > Adjust > Threshold ).Then open the Corsen plug-in dialog box (menu Plugins > Analyse > Corsen Plug-in ). You can enable or disable the following options :

Option Role
Display results At the end of the plug-in execution, a table will show several informations about segmented particles 3D (such as area, volume, sphericity, intensity...).
Exclude on edge All 3D particles that are on edge of the image will be removed from the output file.
Change the output file name A dialogbox to select the name of the output file will appear after the execution of the plug-in.
No confirm save dialog By default a dialogbox ask if the result file must be saved. Enable this option to disable this dialogbox.
Show particles 3D After the end of the execution of the plug-in, a new image stack is created with the particles 3D in different colors.
Pack particles data Use this option when your memory is low to handle large segmentation work. Corsen will reduce the memory footprinting of the particles 3D object.

Once, your options selected, launch the 3D segmentation of Corsen by pressing OK. ImageJ will ask you if you want to process all the image of your stack, answer yes.

When all options are disabled the plug-in will execute the 3D segmentation and will ask you if you want to save the result file. The default name of the output file will be the basename of your image file with a " .par " extension. If your image to process is no linked to a file, a dialog box will appear for selecting the output file.