Help! Thresholding lipid droplets on cells

Hi there, I am new to ImageJ and have been trying to successfully threshold regular colored images as well as red stained images to quantify the fractional area covered by lipid droplets which accumulate on top of my cells. I have experimented with manually setting thresholds and the results are okay, but it is still not as clear cut as I would have hoped. Attached are some images I am trying to analyze. Any help is greatly appreciated. How can I go about use of an automatic thresholder?

Have you tried to segment your images with the Trainable Weka segmentation plugin?


You can also try to use the Morphological Segmentation plugin. These are the result I got after converting the second image to 8-bit and inverting it:


Thank you for the ideas, I will try to download the plug-ins you both mentioned and get back to you! They look promising! I found out that the smaller circles can be ignored so that should make it easier to find an appropriate auto-thresholding method.

1 Like