Hey,
I did a Ki-67 immunohistostaining (green), also a DAPI (blue) staining to make the nucleus visible.
As expected Ki-67 stains where the nucleus is located.
In order to calculate the Ki-67 labeling index, I would like to count the spots where green (Ki-67) and blue (DAPI) overlap.
I tried ImageJ and Imago to no avail, since both are only able to count all cells/nuclei in my taken pictures.
I also had my hands on a very old version of MetaMorph with limited functionality which didn't have the options I was looking for.
In Image J you can count the number of particles for each staining. Would that solve the issue?
Try CellProfiler software (Broad Institute) for image segmentation and analysis. You will find analysis pipelines on their website which will solve you problem.
In ImageJ you can use the green channel to make a mask for the blue channel then only count nuclei in that area. The limitation to this is that if you have cells overlapping each other, any nuclei under or over a cell that has positive Ki67 staining will be counted whether it has Ki67 or not. This should only happen if your cells are in suspension, or freshly settled out at high concentration, or have grown beyond confluence.
First of all, I want to thank everyone for giving helpful advice.
I'm still in the lab and didn't have time to try anything out yet but hopefully will tonight!
@Ron: How do I make masks? I rarely work with ImageJ, so I'm not familiar with it's functions...
@Martin: Thank you for suggesting CellProfiler, I will definitely give it a shot.
@Patrice: Thank you for the awesome manual, I will try it as well if I find the time tonight!
@Pavel: How is the number of particles related to the amount of overlaps? Seems like I don't get the connection...
If you count Ki positive nucleus, lets's say =x, and total nucleus =y, then therotically you may get x/y. Presumably for the info you gave, all Ki must be nuclear right. If this is true, then at the end you do not count overlaps, you just infer them.
Doesn't Ki-67 stain for cell proliferation? I was under the impression that this protein localizes to the nucleus only during proliferation phases and should be absent for resting or apoptotic cells. Shouldn't you use something like the TUNEL assay to get the apoptotic index?
Anyway, If you currently have a composite color image, you should separate the channels using "Image - Color - Split Channels" you can then adjust the the threshold of the channel you want to use as a mask. Once you are happy with the threshold levels (without clicking apply in the threshold dialog) go to "Edit - Selection - Create Mask." Now you should get a new image that looks like the threshold selection. Use this to mask the other channel before counting the nuclei.
I know this is an old thread, I would defintely recommend you visit www.MIPAR.us. It has been able to solve problems for us nothing else could.
It is very powerful and intuitive 2D/3D image analysis software written by scientists/end-users and is about to be released as a free trial on December 14th. You can sign up on the site as well as submit images/datasets to test.
There are many image analysis options available, but I think you may just find MIPAR to be something special if you give it a try!