I have completed the immunofluorescent staining, with antibodies A and B to co-label our intended cells (A+/B+). Then, how should I do to get a percentage or numbers of colabelled cells? From taking immunofluorescent images to image processing software (Image J) application, is there a detailed procedure?
Thanks very much！
it will be tricky to use a image processing software for this.
May be you can take multiple (8-10) 10x images and manually count the number of cells labelled with each or both the antibodies. You can count DAPI for total number of cells
I used Image J for counting B lymphocytes stained with immunofluorescent,and I have good result. Good lucky
Many thanks to Dayma. That is I should take enough images randomized (?) for every slide as soon as possible, and the magnification should be 10× (?), or the vision we can distinguish labelled cells?
Thanks Peralta. Image J is really a wonderful software. I will try it later.
Hi Hailong, try to binarize the images. This will produce black and white images of channel 1 ( antibody A) and channel 2 (antibody B).
Use the "analyze particle" to count how many cells you have in each black and white image.
Then, use the logical "AND" on the black and white images to generate a new black and white image containing only the cells that are positive for both antibody A AND B.
Apply again the " analyze particle" function to this new image to count these double labeled cells.
You will now have the numbers of A+, B+ and A+/B+ cells and will be able to calculate and express these as percentages.
good luck!
Thank Santini for your help. To “analysis the particle” is really a useful method for cell counting, and I have tried it on puncta analysis. But how to setup the “size" for counting to distinguish cell and interference? The original images taked by confocal in 10× or 20× are both ok for calculation?
Without seeing the images it is a hard to know what would work best.
Are the interfering objects debris? Are they smaller than the cells? or are they also dimmer?
Few suggestions are :
1) set up an high treshold before binarize the images so that low intensity (interference) objects will not be "captured";
2) run a "erode" filter on the images until you have removed objects that are smaller than the cells.
3) create an histogram of the area of the objects identified on the images. Set size range - minimum and maximum - so that only the objects that represent your cells are identified.
I hope this help.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication