I want to check the fluorescence intensity of two different proteins across different samples of adenoma. I am using ImageJ for the same. I am first turning the image into 8- bit grayscale image->setting the threshold-> calculating the integrated optical density and area of the fluorescing region-> integrated density/ area to get average optical density

(Found this from a paper). I am not able to find an optimum threshold that can be applied to all the images of different samples. Is this method even correct? After choosing a single threshold for all the images, some region of fluorescence intensity is lost from some images. Are there any other methods to calculate the fluorescence intensity? Also, I do not have any control with which i can compare the samples. What should be done in such a case?

All suggestions are welcome.