Mark yeast cells

Hi everyone,

I’m trying to use cellprofiler to identify some yeast cells and measure the ratio of intensity from the nucleus / cytoplasm. Unfortunately the images are grayscale, so I don’t have different colours for nuclei… The images (e.g. im2.tif which is attached) have also a bad contrast and I can’t identify the cells very well in the first place. Could you please help me with a pipeline which can improve the detection of these cells? Is there a way to measure the intensity from the nucleus/cytoplasm if I have only primary objects (whole cells)?

Another unrelated question that I have: is it possible to integrate the cellprofiler modules in Matlab? I know that the first versions of cellprofiler used Matlab code.

Thanks,
Razvan

Without additional stains, what you are asking is not easy. However, I’m attaching a pipeline which makes an attempt. It finds the foreground, masks the image with it, then tries to identify the nuclei by perform per-object thresholding. The only reason this can work is that the nuclei are fairly bright against the cytoplasm.

Now that CellProfiler has switched to Python, there is no longer a link between CellProfiler and MATLAB, except for the output file. You can call CellProfiler from the command line in MATLAB (just like any other application) by running CellProfiler “headless”, that is with out the GUI. If you are using Windows, type “cellprofiler.exe --help” from the installation folder to see a list of arguments that can be passed to CellProfiler.

Regards,
-Mark
2011_09_02.cp (7.04 KB)

Many thanks,
Razvan