to perform automated counts of fluorescently stained cells.
M. Fero (7/2004)
This protocol describes semi-automated cell counts using fluorescently
labeled cells, a hemocytometer and ImageJ software. The
is not needed if you have already calibrated the areas of images taken
with your microscope and camera. Our setup automatically imports
images in iPhoto and loads them into Photoshop with a
However, you could also import and crop the images directly into ImageJ
if you prefer.
Cells in media
1x P.I. (100 µg propidium iodide / mL PBS)
U.V. fluorescent microscope with camera
Computer with ImageJ and image cropping software.
Fix Cells by diluting them 50% in EtOH.
1. To an eppendorf add 0.1 mL cells in PBS or media
2. Add 0.1 mL of 100% EtOH while vortexing.
3. Spin at 3,000 RPM x1 min.
4. Aspirate off supernatant and respend cells in 0.1 mL of 1x
P.I. Mix, incubate 2 min.
Photograph cells on u.v. microscope
5. Add 10 µL of stained cells to Hemocytometer.
6. Place on U.V. microscope. (If the cells are too crowded then
you may need to further dilute them in P.I. or PBS)
7. Photograph 1 large square of hemocytometer with brightfield.
8. Without moving the stage or changing the zoom, photograph the
same area under u.v. light with red filter.
Import and crop photo (e.g. using
iPhoto and Photoshop)
9. Import photos from the camera to the computer (e.g. with
iPhoto). Double click an image to open image in
10. In Photoshop, use the marquee tool measure the size of large square
of the brightfield hemocytometer image.
(A large square on the hemocytometer is
1mm x 1mm x 0.1mm = 0.1 µL)
11. Crop the u.v. fluorescent image of cells to same size as a large
12. Convert to grayscale (Image > Mode > Grayscale).
13. Save file as .jpg image (medium resolution).
Count cells in cropped image using
14. Open .jpg file with ImageJ software.
15. Convert to binary image (Process > Binary > Threshold)
16. Count cells (Analyze > Analyzed Particles... check Display
results, Clear results table, Summarize).
17. Repeat counts with additional large squares to be sure that the
cells are evenly distributed and to minimize stochastic error.
(The standard deviation is ~ sqrt(mean)).
Note: An alternative to cropping the image is to instead photograph
the fluorescent cells with a low level of brightfield illuminescense in
order to visualize the lines of the hemocytometer together with the
fluorescent cells. Select the area of the large square for cell
counts directly in ImageJ.
Calculate the cell density
Cell density (cells/mL) = Average Cell Count (cells/large square)
* 10,000 (large squares/mL) * d.f. (dilution factor, if any)