Cell Detections for Binary Image

Hey everyone,

I’m trying to use QuPath to detect all the segmented nuclei from a binary mask so that I can extract their features later. I’ve tried to get maximal detections by playing around with the parameters but it’s always missing a significant amount. This is a binary channel, so shouldn’t there be a simple way QuPath can detect all the white regions or the image?

You might be better off trying SLICs or DoG segmentation rather than nuclear, but for the moment it looks like most of the missed objects are the largest objects, so much of it might be a size threshold issue.

I’d suggest using analyse particles in ImageJ and then sending the overlay to QuPath.