Is this for measuring DAB or other chromagen density in your sections? Are you after a standard density measurement or would a true positive object count be more useful? Or maybe both combined. Also the formula you have listed will give you the lower value, changing the first minus to a plus will give yiu the upper value.

Dear Dr. Alkathaib. We work using serial sections immunostained for c -fos. Our experimental paradigm is designe to analyze the olivary complex after sound stimulation in order to elucidate the effect of auditory cortex ablation. We want to compare at different survival times: number of immunoreactive neurons, the density of the immunoreaction and the perimeter of the particles. Our cromogen is DAB with Ni.

Dear Dr. Nowell. The lowest value is what I am looking for, in order to avoid segmenting neurons near the background value. Do not quite understand what you comment me the sign in the equation. Thanks a lot for your comments.

Ahh ok, the reason for asking about the formula is that any threshold has to have a lower and upper value. the formula you posted will pick the lower value, but you still need something for the upper value. in your case that may just be the maximum of the image intensities.

The formula you posted is a simple auto threshold method (there are many of these).

Counting DAB stained cells in immunosections is fairly straight forward. We routinely use the colour deconvolution system built in to the free Fiji distribution of imageJ (availible here www.fiji.sc). With this you can split the haemotoxylin counterstain and the DAB stain into two seperate channels and then use those for further analysis (area coverage, intenisities, object counts, object descriptors etc).

Fiji incorparates 16 differnt autothresholds out of the box and you will be able to find one that suits your needs.

If you can, post up an example image and i can work out the protocol that works best for you. Then you can either follow the steps or if you have lots to do i can write up an automated script for you to be able to batch process many images in a folder.

Thhanks a lot

this is a peace of a mosaic taken with x10 with a motorized stage. The reaction product is black.

Means and SD of samples form an experimental group

How to quantify milk on the lumen of alveolar secretory cells on mammary gland using ImageJ?

Please send one of your image and I will give you different options.Regards

One is really impairing the counting of Fos positive nuclei when you use a colored counterstain that cannot be eliminated by filtration. For example instead of using Cresyl violet, or Thionin, or hematoxylin etc, one uses Neutral Red or Methyl Green, then placing an appropriate Red or Green filter into the system totally erases the Fos-negative cells and you can very easily count the weak or strong Fos (especially if you adjust your conditions to having NO background staining). If you are getting background staining, your assay conditions should be adjusted appropriately as is outlined in my article on Titration of antibodies. To then get total cell counts (with or without Fos) you put the complementary filter in (red counterstain; green filter and green counterstain, red filter). Then you do not have to worry about the segmentation of weak Fos being similar in gray level to that of the counterstain.

Hi Miguel,

here is the result the auto threshold in Fiji/image gives. The values it calculates are pretty much the same as the ones your formulal generates. The Fiji one detects a few more dark pixels, which is probably a good idea for picking up a dark stain. There really isn;t the resolution in the image you posted to detect individual cells. Do you have a higher res one you could post?

At the moment the best we can get i area coverage and instenisy/desnisty statistics.

Thanks Dr. Hoffman and Nowell for your advice. I will try to modify our approach with your useful comments.

Warm regards.

M. Merchan

no worries, anytime. If you get stuck and need a bit of scripting done just let me know