How to deal with background signal on a channel on CellProfiler 3d monolayer analysis

Hello,
I have altered this pipeline a bit for my use.


I need to count number of nuclei (DAPI) and how many of those nuclei have GFP+ cytoplasm as well as their volume. I’m having an issue where due to high background signal in DAPI for some images, or possible bleed through from other channels the nuclei aren’t being segmented or counted correctly I have attached examples out the outlines below. I tried to add a threshold to remove large segments but that didn’t work, instead it just counted much less nuclei. What can else can I do?
Thank you so much in advance

my pipeline tumor analysis2.cpproj (1.0 MB)

ex image3d_1wkAPCpsesg2_xy1_ch1-2.tif (6.0 MB) 3d_1wkAPCpsesg2_xy1_ch2-2.tif (6.0 MB)

how its segmented3d_1wkAPCpsesg2_xy1_ch1-2nuclei-outlines.tiff (2.3 MB)

post max volume filter
3d_1wkAPCpsesg2_xy1_ch1-2nucleioutlinepostfilter.tiff (777.9 KB)

Hi @mquintero,

You’re off to a good start! I would suggest changing your threshold (lower and upper bounds). You can hover over the pixels on your image and find their intensities. Then based on those intensities, define the limits of your threshold. You would then get a better segmentation.

2 Likes

Hi @mquintero,

As mentioned changing your thresholding is doing a good job in your pipeline. Also just tried with smoothing a bit helped much better. PFA pipeline with the modified thresholding values.

tumor analysis2_new.cpproj (1004.8 KB)

Regards,
Lakshmi
Fujifilm Wako Automation (Consultant)
www.wakoautomation.com
For CellProfiler training or optimised pipeline write to,
lakshmi.balasubramanian.contractor@fujifilm.com

Read more on our site.
Yokogawa CV8000 - The Ultimate in Confocal HCS
https://www.wakoautomation.com/products/yokogawa-high-content-imaging

Join us at, https://www.slas2020.org