Urgent ERROR!


#1

Dear Profiler Gods,

I am having an URGENT error trying to analyze cells using the Identify Secondary module:

There was a problem running the analysis module IdentifySecondary which is number 06. Error using ==> CPretrieveimage at 57
Image processing was canceled in the IdentifySecondary module because CellProfiler could not find the input image. CellProfiler expected to find an image named “SmallRemovedSegmentedCells”, but that image has not been created by the pipeline. Please adjust your pipeline to produce the image “SmallRemovedSegmentedCells” prior to this IdentifySecondary module.

Stack:
CPretrieveimage in C:\Documents and Settings\Administrator\Desktop\CellProfiler_1.0.5122_XP64\CellProfiler_1.0.5122_XP64\CellProfiler_mcr\CPsubfunctions\CPretrieveimage.m (57)
IdentifySecondary in C:\Documents and Settings\Administrator\Desktop\CellProfiler_1.0.5122_XP64\CellProfiler_1.0.5122_XP64\CellProfiler_mcr\Modules\IdentifySecondary.m (271)
AnalyzeImagesButton_Callback in C:\Documents and Settings\Administrator\Desktop\CellProfiler_1.0.5122_XP64\CellProfiler_1.0.5122_XP64\CellProfiler_mcr\CellProfiler\CellProfiler.m (9409)
gui_mainfcn in C:\Documents and Settings\Administrator\Desktop\CellProfiler_1.0.5122_XP64\CellProfiler_1.0.5122_XP64\CellProfiler_mcr\CellProfiler\CellProfiler.m (11138)
CellProfiler in C:\Documents and Settings\Administrator\Desktop\CellProfiler_1.0.5122_XP64\CellProfiler_1.0.5122_XP64\CellProfiler_mcr\CellProfiler\CellProfiler.m (55)

I also have a link to the pipeline and a sample image:

homepage.mac.com/jzi1/FileSharing24.html

Thanks a million !!!

-CellProfiler Dummy


#2

Hi Kevin,

Change the drop-down at the top of the IdentifySecondary module (“What did you call the primary objects…”) to be “Nuclei”, and I think everything will work for you.

Ray Jones


#3

Thank you for your reply Ray.

Now that I am over that hurdle, I was somewhat disappointed by the result of the IdentifySecondary module. I am working with multinucleated cells, and I cannot seems to tweak the settings in such a manner that cells are correctly defined.

Do you have any tips or suggestions?


#4

Hi Kevin,

It all depends on your images. However, the link that you gave earlier is dead. Can you repost a couple images? You should be able to post them here directly in the forum if that’s convenient.

David


#5

David,

I have reposted the images.

Thank you for your help, I really appreciate it.


#6

Hi Kevin,

I retrieved your files from homepage.mac.com/jzi1/FileSharing25.html and not the link that you previously posted in the forum. My comments are based on the two files I found there: B3_num2.tif and HelpPipeline.mat

The effectiveness of the IdentifySecondary module is strongly affected by how well IdentifyPrimAutomatic does. If the primary object identification is poor, then the secondary object identification won’t do well either.

From your files, I see that the primary objects in IdentifyPrimAutomatic (defined as Nuclei) are being split into pieces in places which it looks like there should just one object. Since these objects are being broken up, the smaller pieces are falling outside the diameter limits you set (min: 9, max: 17) and you are getting only fragments of cells. I’m assuming here is this probably not what you want.

These fragments are then being given as input into IdentifySecondary, which doesn’t do much better. Perhaps tweaking the parameters in IdentifyPrimAutomatic would be a good place to start, to make sure your objects are being identified the way you want to begin with.

Afterwards, you can run IdentifySecondary in test mode (look at the last setting in the list for that module) to see the performance of each of the different methods for defining secondary objects.

Regards,
-Mark