At one time we tried a similar method, but came to the conclusion that direct human counting is better (there are significantly fewer errors). If you do not have enough time, you can first photograph 100 or 200 random fields of vision and later count from the photograph.

(PDF) The RhoA-ROCK1/ROCK2 Pathway Exacerbates Inflammatory Signaling in Immortalized and Primary Microglia

See section 2.6 for how to count the DAPI stained cells with imageJ. You can also potentially do this for your marker of interest depending on localization.

Definitely!

You need to capture images of just DAPI and just the marker's signal. Then analyze using ImageJ to get the area of each signal, and then you calculate the ratio of the marker signal area over the DAPI signal area.

These resources helped me a lot:

https://www.youtube.com/watch?v=mqpo1olsT7Q

https://www.youtube.com/watch?v=8c68qIz_ftw

https://imagej.net/ij/docs/menus/analyze.html#:~:text=Area%20Fraction%20%2D%20The%20percentage%20of,in%20the%20stack%20or%20hyperstack