Hi, My PhD student is stuck with COMSTANT and I ran out of ideas, can someone help us identify where the mistake is? Thanks, below is hisquesiton.

I am trying to quantify bacterial biofilm by live/dead kit (green/red stain).  I am using a Leica Confocal Microscope and COMSTAT plugin in ImageJ to analyze the images. Although I follow the COMSTAT manual, I do not get correct quantification of the biomass. I see that, after generating the OME-Tiff stack, COMSTAT quantifies only pixels of some of the images of the stack and not all pixels of all images (or at least that what appears to me) resulting in a very low biomass.  I do not understand where the mistake comes from, for example, is it from specifying image properties in the ImagJ? Or from choosing the color channel (8-bit, 16-bit…etc) or from other step during the process I am not aware of??  I have not changed the default threshold setting in ImagJ.  Here is what I do:

My confocal microscope gives me the captures in individual images TIFF and I have to convert to OME-stack by ImageJ. The image information from the Leica confocal microscope is:

1024x1024 pixels

 voxel-width 240.5 nm

 voxel-height 240.5 nm

voxel-depth 377.7 nm

step-size 0.38um

1) The first step that I follow is: File>Import>Image sequence> load all individual imagen to get the stack file. This stack has just 1 channel (I just select red or green images).

2) I change image properties modifying unit of length (inch into micron), pixel width, pixel height and voxel depth by filling in the information from image information file produced by the microscope. I understand that pixel size in ImagJ is the voxel size in Leica file…or is this wrong?

3) I export the stack by bio-format (plugin>bio-format>bio-format exporter) to convert into OME-Tiff in a new folder (here I click uncompressed).

4) Finally, I run COMSTAT, add the OME-Tiff image that I created, click biomass and thickness distribution and go.

The results I get is very low biomass and that do not represent what I visually see in the images.

 Analyzing visually the images and comparing what I see to the Biomass calculated by COMSTAT, I understand that the program counts only some pixels but not all, or may be counts the pixels of some images but not of all images.