WatershedCounting3D

[Simon-P]

Hello,
I would like to find a solution to do automated counting of particle (puncta staining of synapse) in a stack of image obtained with a confocal. I have try the plug-in "3D object counting" but as my background is not very homogeneous the result is not reproducible. I would like to use the "WatershedCounting3D" plug-in (Thomas J. Gniadek and Graham Warren) because I think it should solve that problem as it does not use threshold. Unfortunately when I try it, the object found are not accurate. As you can see on the image it found object around the real particle.
I think I have problem with the setting. The one used for the image posted is:
x=0.207; y=0.207; z=03; max radius search=0.5; Local Maxima sampling dimension: x=2; y=2; z=1
I read the paper describing the technique (http://www.ncbi.nlm.nih.gov/pubmed/17319897) but I don't find any explanation on how to choose the best setting. I will greatly appreciate any advice,
Thank you very much.

Simon Pieraut

Watershed3D_Results_ImagePlus-2.tif

Watershed3D_Results_ImagePlus-1.tif

[Simon-P]

Hello,

I also wondering if there is a way to first make a 3D ROI with my stack and then count the particle using WatershedCounting3D only in that 3D ROI. My goal is to count number of puncta per volume of tissue. As my puncta are only present in a part of my image, I would like to only count in that area.

Thanks,
Simon.