We are tracking cells using the FUCCI marker and we try to deduce morphological characteristics about the nuclei. At some point the segmentation of the nuclei is not good enough and we get lumps form in the image that do not allow a good evaluation of these properties. At first the cells are well identified and the problems start when the density of the nuclei in the images increases. Attached frames in which the nuclear density is high and the identification is not optimal.
Thanks a lot!!
WhatsApp Image 2021-04-18 at 13.06.34|690x362
Final_Update-9.4.2021.cpproj (805.6 KB)
You mentioned that it works for few cases & problem is only when number of nuclei increases. I checked your pipeline. It seems to be fine. It would be helpful if you could share one of your raw image to reproduce your error & so we can help you better. Nevertheless, here few suggestions you can try with the test mode,
- Try the shape option in the method of drawing lines.
- Click “Yes” in the Log transform option
& the other would be the playing around with the declumping parameters (enter manually)
Hope this helps.