Looking to Count Goblet Cells in S. Intestine section

Hi all!!

I am trying to count goblet cells in a gut section that was stained with H&E using Fiji ImageJ. Essentially trying to make a macro that can easily count the cells that I have multiple images for. (90 +)

I played with the thresholds but could not find the best way to do it.

Any Ideas!?
Goblet cells are circled in red.


Thanks in advance :grinning:

Jon

@jhd8593

Best way for us to help you is if you share an original image file here. That way - we can actually see your images. And please indicate what you consider a goblet cell (I actually know what they are… but a lot of other folks on here might not. We are all coming from different scientific backgrounds.)

1 Like

Can you also share an unedited, non-labeled, original image file here with us?

To start… you can take a look at Trainable Weka Segmentation in Fiji and also QuPath. Both might provide the best solutions in this case… but having the image file will help us test things a bit too.

2 Likes

Looks very similar to


unfortunately, which I also wasn’t able to get to work well. At least in QuPath.

Weka has more texture based options, I think, and might be better if your files are small enough to open in FIJI. Otherwise, you could definitely use QuPath to chop your file down into FIJI sized sections, and then use Weka on those.

1 Like

Definitely got Weka to work and analyze my cells. :grinning:
Going to look into Qupath and see what I can figure out from there.

Weka worked perfectly! Thank you so much! Will look into Qupath.

3 Likes

Hey Jon.

I’m a bit late to the party but I hope you can help me out as well. How did you get Weka to work on the quantification of GC? Would you mind giving me a shor explanations of the steps you took? I’d really appreciate it