Analyzing ImageStream cytometry data with CellProfiler

Dear Sir or Madam,

I am writing to ask for your advice about the problems our laboratory encountered during the processing of the ImageStream data with CellProfiler.

First question. This article provided us an example of user-friendly solution for working with CellProfiler, when .cif file is uploaded directly to CellProfiler. Unfortunately, we could`t do this with ImageStream .cif file Could you explain how we can make it?

Second question. Now we process Image Stream data after image extraction from .cif file. Our purpose is to count black dots on cell’s surface. Unfortunately, function IdentifyPrimaryObjects in CellProfiler allows to count only light objects on dark background. Threshold in CellProfiler does not allow me to distinguish black dots. Due to this I have to invert colors manually with python script. Is there in CellProfiler possibility to invert colors in image or count dark objects on light background?

Third question. In some articles and on YouTube I saw examples of image classification using machine learning with CellProfiler Analyst. Could you tell me please is CellProfiler Analyst able to count black dots on cells using machine learning? Could you explain me how to do this task if it is possible?

I use CellProfiler 3.1.5.

I look forward to your reply.
5Au%20OL_9__375_Ch1 5Au%20OL_9__375_Ch1
Best regards,
Ivan Kotov, laboratory of nanobiotechnology, MIPT

Hi Ivan,

I’m pretty new to CellProfiler, but to answer Question #2 I believe CellProfiler’s “ImageMath” Module has an “invert” Operation.

Sorry I can’t help with your other questions.

Rob
FUJIFILM Wako Automation
https://wakoautmation.com

1 Like

Yes +1 on @RobBukar you can use the ImageMath function. Not sure whether invert is built in, but you can achieve it with ImageMath.
If your dots are always that much brighter than the cell and background you can simply segment them using the IdentifyPrimaryObjects modules, no need for machine learning.

Hi,

The .CIF you’re using, is it a gated population? i.e. you may need to do first some preliminary gating on IDEAS software, and then hit “Tools > Create population from gate…”

It will then export a .CIF file that contains the cells. Only this type of .CIF can be used for our pipeline.

The raw .CIF without exporting will be not be useful for the pipeline.

Hope that’ll help.

1 Like

Hello, I am also trying to import a .CIF file into CellProfiler (latest version 3.1.9), and did already limit the population in IDEAS (version 6.2) as you mentioned (the resulting CIF file is about 1.3 GB), but it is not loaded in as an image file. Am I missing something about the drag-and-drop procedure for CIF files? I simply installed CellProfiler and prepared my CIF data as suggested, and dragged-and-dropped the file in. It shows up in the file list but cannot be used in a pipeline, I get the error “The pipeline did not identify any image sets”. Also, if I double-click the CIF file in the file list, I get the error “Failed to open file, <path_to_file>”. The pipeline I am trying to run in test mode is the sample pipeline available at https://cellprofiler.org/imagingflowcytometry/.

That link, BTW, does not make any mention of the drag-and-drop functionality described in the Hennig paper, not sure what to make of that.

Thanks!