Problems with Identifying Primary Objects

identifyprimobjects

Background

I am looking for a fungal disease on bat wings.

Analysis goals

I need to find the percentage of wing that is fluorescing orange.

Challenges

I have issues when identifying primary objects to find the orange section of the wing. I have successfully gotten it to show black when I unmix colors, however I don’t know what setting will allow me to select it with identifyprimarypbject. It seems to select all areas but the orange part of the wing. I have messed around with thresholding and size of object to no avail.

I have included my project and images via this google drive link . Please let me know if you have trouble accessing. Thanks for any help. :slight_smile:

Hi @robobrooke,

If you’re only looking for the percentage of the wing which is orange then it might not be necessary to try to identify individual areas as objects. What you might want to do is use the Threshold module to create two binary images - one representing the whole wing and another featuring just the staining that you wish to consider as positive. You can then use the MeasureImageAreaOccupied module in “Binary Image” mode to calculate the area of each detection. The CalculateMath module could then be used to run Area_Orange / Area_Wing to generate the score you desire.

However, it does look like your staining is represented as dark regions rather than bright ones. CellProfiler generally ‘sees’ light regions as positive, so you might need to either invert the images using the ImageMath module or just subtract the original results from 1 (or 100 if you made it a percentage).

Hope that helps

Yes, I was thinking that would be a different way to do it. Thanks for being clear about calculating the area in that way.