I am extremely new to both CellProfiler and ImageJ and was hoping to measure the fluorescence intensity or pixel intensity of the cell membrane of individual cells. What is the best way is the best way to go about this? In previous papers I have read, certain people have been able to trace the cell membrane in ImageJ with the freehand tool and measure the intensity of the perimeter in ImageJ. However, I am not sure how they were able to do this because when I use the freehand tool, it appears that it is measuring the area within the drawn lines - not measuring the intensity of the pixels of the perimeter.
Currently, I am working with yeast cells stained with Filipin III (fluorescent under the DAPI channel) that has been localized to the plasma membrane. I have attached an example of a tif image of the type of images I am currently working with.
The most likely way they did that was to use the Freehand tool followed by Edit > Selection > Area to Line followed by Analyze > Plot Profile. You’ll need to average these values (which are the intensity for each pixel along the perimeter)
Ideally we would try to do these measurements automatically
CellProfiler provides a bunch of Perimeter* metrics, but you’ll have to be careful when detecting the individual cells automatically…
The detection of edges of the cell will depend on the intensity of the signal of the cell. And you have some highly verying intensities…
If you want to automate it, you may need to use a secondary marker that does not vary in intensity. and staing the whole yeast. This can then be used to first find the cells automatically, and then use the Fiilipin III channel only to measure that intensity metric you want.
Now onto imaging.
The other thing is that microscopy can be hard to get absolute intensity values. I’d suggest looking into what you can use to normalize your data with an internal control, like for instance, the intensity at the center of each yeast, if that makes sense.
Problem is, intensities are dependent on sample, lamp age, exposure time, extra light around the microscope, etc… and can be hard to relate an intensity to a real concentration. So to use these measurements in absolutes, you should probably create a standards of known Filipin III concentrations so you can build a calibration curve of ìntensity vs concentration.
It depends on the ranges of intesities you expect to have, but I notice that your images show uneven illumination (which is usual for microscopy images). It’s not much, but there is roughly 10 gray levels of difference between the background at the edge of the image versus the background in the center of the image. There are two things you can do.
Acquire images of a Fiilipin III sample (just the dye) to get the field illumination structure and perform flatfield correction
Crop the image to minimize the differences in background illumination.