Try my attached pipeline. I think you are very close, but here are some key issues:
(1) ApplyThreshold - I removed as it wasn’t doing anything. I assume you were just testing it, but it doesn’t add much beyond IdentifyPrimaryObjects
(2) IdentifyPrimaryObjects – I disabled your version, but kept it as a reference. My IdentifyPrimaryObjects has a few changes:
- Global Otsu thresholding. Avoid Manual thresholds as they won’t be robust across samples, especially for histology
- Three class thresholding helps when there are background, lighter staining, and darker staining. Since you want to only include the darker staining (to my eye), then you can set the middle class to be included with the “Background”
- Threshold Correction Factor. This is your main adjustment for thresholding here. Please adjust accordingly – Test across other images. I set this empirically to my eye, but you will almost certainly want to adjust this.
The rest of the settings are fine if you only care about measuring overall intensity of the dark DAB regions.
DLpipeline.cppipe (11.8 KB)