Creating a simple mask (Cell Profiler)

Sample image and/or macro code

Bat Wing

I used a polygon in ImageJ to select the area of interest (a bat wing) and now I want to move the image over to cell profiler for analysis.

Background

I am trying to get a mask of the wing area so I can identify fluorescent areas on the wing. I prefer to use CellProfiler for this because I already have a pipeline to use for the future analysis.

Analysis goals

Do I use identify primary object? What parameters do I set so I get one object (the wing)?

Challenges

When I use the identify primary object it breaks down the wing into 5 or 6 objects instead of the one (wing) object.

Hello Brooke,

Thanks for the detailed information.
Yes, you can use IdentifyPrimaryObject. You just need to adjust the size.
You can measure length from Tools -> Measure Length to find the size.
For the image you provided:


To learn more about these modules, take a look here.

Hope this helps!

Do you have any advice on threshold? This is the closest I have got…but it is still selecting some areas that aren’t orange (my ROI).

uvzoom pdzoom

I’m assuming you’re talking about the triangle shape part that is selected, correct?
You could shrink the object before masking and then select your regions of interest.
For thresholding, it’s hard to tell without seeing the pipeline. You could try otsu and define a region as part of background, but it depends on the intensity of that region compare to the ROIs.

Here is a link to my photos and pipeline. I don’t know if you would be able to look at it, but I would greatly appreciate it if you did.

https://drive.google.com/drive/u/0/folders/1vEnRR-Ie_ZoR9RYpKEHA2LCRSlVCwQOw

Hi @robobrooke,
In your pipeline, you may have to add a “ImageMath” to invert the image. Once the image is inverted, then the segmentation looks fine. PFA output & pipeline.

new_test_pd1.cpproj (408.8 KB)

Hope this helps.

Regards,
Lakshmi
www.wakoautomation.com

Thanks Lakshmi! I will check this out.