I'm trying to measure the colocalization of two proteins in a .tiff image taken with a wide-field fluorescence microscope. In ImageJ I have split the image into the two channels (red and green), and now I have to apply a threshold, however when I threshold both channels the bright areas corresponding to the cells are located in the darkest part of the histogram instead of the brightest part, as you can see in this image for the red channel:


Why ImageJ is assigning darker values to the cells rather than to the background? I'm having problems with this as I cannot properly select these cells for the colocalization analysis. Anyone does have any clue? Any help would be greatly appreciated.


Check the "dark background" checkbox that is unchecked if the background is dark and you want to select the bright cells. Right now it is assuming the background is bright.

  • $\begingroup$ I was thinking about that too but it just inverts the colors, simply assigning white to the cells while they are still considered the darkest objects in the spectrum: screencast.com/t/INdZ5pFLHgIn Here instead how cells (or metal particles like in this case) should appear on an histogram: screencast.com/t/TlFxRwsXubr $\endgroup$ – betelgeuse Aug 6 '19 at 5:52
  • $\begingroup$ On the last picture, the particles are bright so they appear as a peak on the far right of the histogram, while the other peak on the left represents other objects in the background that have been excluded by thresholding them out. You may notice on the histogram that this left peak is inside a rectangular black frame representing the background. And as you can see the “Dark background” box is not checked. With my image I get exactly the opposite, which is weird… and I don’t know what am I doing wrong. $\endgroup$ – betelgeuse Aug 6 '19 at 5:53
  • $\begingroup$ @betelgeuse That's from your actual data then and can't be fixed with imageJ. Look at your histogram vs that one: it's doing the best it can do with what you gave it. $\endgroup$ – Bryan Krause Aug 6 '19 at 5:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.