Particle analyses finding cell count in a picture

hi all
I am new to this software and I tried following few videos online
I have an image that I would like to do analyze particle to get cell count

I did 8-bit conversion
subtract back ground
adjust threshold
fill holes
convert to mask
watershed

but the result is far off

I attached the image

https://www.dropbox.com/s/zlz1es0bteb3n6s/%233BaselineB-01.tif?dl=0

if someone could walk me through the instruction I will be grateful since I have 1000s of these images

@khaled_cs

If I were you… I would focus on segmenting and counting the nuclei. There is no way for you to do otherwise (too much overlapping of cells and difficult delineations).

So I would do something like:

  1. Open via Bio-Formats
  2. Split Channels
  3. Use the Image Calculator to SUBTRACT C=2 from C=1 (make sure to create 32-bit float image)… It should look like this:
  4. Convert to 8-bit (Image > Type > 8-bit)
  5. Apply Maximum filter with radius=5.
  6. Image > Adjust > Threshold… use MaxEntropy and Auto to then Create Mask
  7. run some morphological operations… such as Erode (helps get rid of tiny stuff)… then run Watershed. Should look like:

    (it’s not perfect… but a start. you can play with more morphological operations to refine things a bit)
  8. then run Analyze Particles… - setting as such:

    (again - you need to play a bit here with the settings…)

All of this is Scriptable - so you can run it as a batch… for more on that - check out this older post:

Hope this helps give you a start at least!

Hi Folks,
Very interesting images. I am not a Biologist, but I was wondering if you noticed what was in the Blue channel?

khaled_cs #3BaselineB-01 Stack2.zip (8.2 MB)
I was just curious.
Bob