In QuPath, you first have to create a Macro by training a pixel classifier. It doesn´t have to be complicated, simply mark the total area of tissue you want to quantify and annotate several examples of positive staining. Do this to train your classifier until you are happy with the result (you can also change the resolution of your classifier) by checking with the Live Prediction tool. It´s important that once you are happy, you should check the classifier on several other samples to make sure it works on all your samples. When you are done, you can load the pixel classifier and measure positivity in your samples.

Hello Magdalena Palic . With ImageJ, you can do Color Deconvolution. If this helps, here's a tutorial video on this: https://youtu.be/5rQl7ejuLSE

Dear Magdalena Palic,

Please read the "Suggestions and limitations" on Gabriel Landini's website (https://blog.bham.ac.uk/intellimic/g-landini-software/colour-deconvolution-2/); you cannot use DAB intensity for protein measurements, you can only calculate number of positive cells.

There are some other plugins in ImageJ that might be of use like https://imagej.nih.gov/ij/plugins/ihc-toolbox/index.html and http://bigwww.epfl.ch/sage/soft/colorsegmentation/