Different Area When Thresholding


I am using the colour deconvolution plugin for my masson’s trichrome stained slides. Here are my steps:
-I use the masson’s trichrome option
-set scale to um using a reference image and checked global
-circle the area of the image I want to threshold
-measure that area
-threshold using one of the algorithms provided
-measure that area again

Then, I get a different area.

Why does thresholding produce a different area before thresholding?


Can I suggest that to understand your query better, you post a recorded macro that shows what are the steps you follow to get that discrepancy. Thanks.


Could it be that in the Analyze>Set Measurements… you have the box ticked at “Limit to threshold”? Then this is intended behaviour.


Yes, I have ticked “limit to threshold.” I don’t understand this feature, could you explain further?

For my purposes, I am trying to measure the amount of collagen/muscle on a cross-section.


‘Limit to threshold’ allows you to measure within a Region Of Interest, but neglect areas that have a certain grey value range, e.g. the areas that do not contain tissue.

In your case, you might want to measure the entire area occupied by the muscle, by drawing a ROI around the muscle (fibre?bundle?). Then you want to measure the area occupied by collagen, so you Image>Adjust>Threshold such that only the collagen is covered by the red color of the thresholding. Now you measure again and obtain another value. Divide the two values to obtain the relative muscle surface area covered by collagen.


Thanks for your response and suggestion.

Sorry for the confusion. Currently, I am measuring the collagen and muscle separately. I have used the color deconvolution plugin which separates the image into blue (collagen), red (muscle), and green. Next, I draw an area around the cross-section for the blue and red images and threshold until blue is not visible in the blue image and the muscles are covered in the red image. Would this be a valid method to measure the collagen and muscle separately? What I am looking for is % of collagen and % of muscle on one cross-section.