What's the best way to quantify fluorescent IHC images?

Hi everyone!

I’m very new to using ImageJ and am about to start quantifying a massive amount of fluorescent IHC images, so I wanted to ask what the best way to do this would be.
The way I understand I should do it is to make a 8 bit image and threshold it, then measure the area.
The thing I’m uncertain about is how to get my thresholds to be consistent in all of my images. I want to use the automated thresholds, but sometimes I need to manually readjust it, so is there any way to make sure the thresholds are standardised?

@kleio

To get started with some general things in ImageJ/Fiji - start with the ImageJ wiki:

You can read up on Segmentation here. There is also a workshop on Segmenting with Fiji. This should be a good place to get you started.

In general, yes, you want to only apply automatic thresholding to ensure reproducibility. Perhaps there are some additional pre-processing steps that you would need to carry out in order to avoid that ‘readjustment’ as you say.

Start with those links above - and then post an example dataset here. Also - specify what exactly it is that you want to quantify. Area? Intensity? etc. That will also influence your analysis workflow.

Hope this helps get you started!

eta :slight_smile:

1 Like

Hi, I’m developing a IHC score method using threshold and color histogram bars
If you want have look this is the video link:

1 Like