Cell counting with ImageJ

Hi, I was trying to do cell counting analysis on some of my cultured rat astrocytes whose boundaries are quite indistinct as the cells are flat and confluent, by watershed after having done some adjustment to the images. The main problem I think I have here is that the colour intensity of the “background” is almost the same as that in the “cell body” on the images. How should I approach it?

1 Like

Hi @Katie_Lui,

would you mind sharing an example image? We could then suggest a strategy most effectively :wink:

Cheers,
Robert

1 Like

Hi Robert,

thank you for your reply!

such as this.

BW, Katie

Dear Katie,
You can use machine learning to segment your image. With Trainable Weka Segmentation (in Fiji: Plugins > Segmentation > Trainable Weka Segmentation

Please don’t use .jpeg files

Analysis with weka image.sc

Citation
In the settings, I had added the Variance and Structure features, increased the maximum sigma to 32, and checked Homogenize classes , see also the screenshot:

Hope that helps.

3 Likes

Hi Alex,

Thank you so much for your reply! I seem to be able to at least get a relatively accurate representation of the area that my rat astrocytes were occupying using Trainable Weka Segmentation.

Follow-up question:
Is there a way to measure the black area on this image? (it is a tiff image but I only attached a screenshot here.)

Thank you!
Katie

Also, is there a way to do the segmentation action on a large scale, i.e. with multiple images at one time.

Thank you.
Katie

I’m a bit confuse if you are able to measure the white area, the black area = total area - white area.

This is some link fr your macro

[Use trainable weka segmentation through macro]

https://imagej.net/Scripting_the_Trainable_Weka_Segmentation

https://imagej.net/Trainable_Weka_Segmentation#Apply_classifier