Apoptosis-TUNEL measurements


#1

Hello,

I have just entered the Cellprofiler community and I am getting familiar with the Cellprofiler.
I would like to count DAPI-stained cells (nuclei) from the confocal images (see attached) obtained from frozen heart sections by using threshold, object size and circularity (shape) criteria. A simple task for an expert, you may say… Could anyone have a look at the attached pipeline and perhaps guide me step by step to set the correct settings.
A similar problem was posted by shupadh, but I couldn’ make it work with my images…

thank you!

best regards,
simon



TUNELPIPE_simon.mat (939 Bytes)


#2

Hi Simon,

I created a short pipeline that loads in your images, converts them to grayscale (most CellProfiler modules do not work on color images), identifies the nuclei, and measures the size and shape.

-Kate
TUNELPIPE_simon.mat (899 Bytes)


#3

Hi Kate,

thanks a lot for the fast reply and your help.
I tried to download the pipeline you wrote, but the downloading of the file recognizes it as a link to the Microsoft Access-Table(?) and I can not copy it in the module folder (txt.files). Please help me get out of here…

thank you,

regards,
simon


#4

Hi Simon,

How are you trying to download the file? You should be able to right-click on the attachment link, select “Save as…” and save it directly.

At that point, you probably should not save it to the Modules folder, but elsewhere on your hard drive. Not because it won’t work, but because Modules directory should be reserved for the installed modules only.

Regards,
-Mark


#5

Hi, Mark

when trying to download the file, I can’t save it as .mat or .txt. file (the format as the module should be??!) elsewhere on the hard disc, but only as a link to Microsoft Access Table, which at the end I cannot open and run as a pipeline (module). The cellprofiler pipeline window displays “loading…” and it gets stuck. Strange??!

regards,
Simon


#6

If you are using a PC, then you should probably get rid of the reference to the Microsoft Access Table extension on your system. Try the following:

  • Double-click “My computer” from the desktop or Start Menu

  • Go to Tools > Folder options…
    from the menu bar. Select the File Types tab from the window that opens.

  • Find “MAT” under the Extensions column. If it’s there, probably it mentions “Microsoft Access Table” under the file types column, or something similar. If so, delete the entry by clicking the Delete button.

  • Close the windows, and try downloading it from the attachment link again.

Regards,
-Mark


#7

Hi, Mark

thank you for the suggestion. It works!

regards,
Simon


#8

Hi,
I have to identify and calculate apoptotic cells. these cells are recogized by 3 features: circular in shape, non overlapping and are have more grey level in their edges as compared to rest of the cells.
First problem that i face is that i am getting a lot of background noice when i run module identify primary automatic. How should i improve image quality.
How should i classify objects on basis of shape and intensity.
I have attached an example image if you need any other information i can provide it.

Regards,



#9

Hi,

We have received questions on using CellProfiler on DIC images and similar images from enough people, that we have posted some suggestions as an FAQ on the forum. Take a look at this page and see if it helps you.

Regards,
-Mark


#10

Hi,
Thanx Mark it was really helpful.
I have another question, in an image if their are objects of different shapes but i want to select only circular objects, is their any way can i do this automatically.

Obaid


#11

One way to do this is to use MeasureObjectAreaShape on the objects, and then use FilterByObjectMeasurement to get rid of the non-circular ones using FormFactor as the measurement. FormFactor is a value calculated as 4piArea/Perimeter^2, and equals 1 for a perfectly circular object. You may want to play around with the cutoff values to make sure you capture as many near-circular objects as you want.

Regards,
-Mark


#12

Hi Guys… I am new user to cell profiler.

I am trying to quantify the fluorescent signals I got after TUNEL staining (green signals).
I used the posted pipeline “TUNELPIPE_simon” to do so.
I have a doubt about the background noise in this pipeline the "typical diameter of objects, in pixels units is (min,max : 2,100) ".
whereas my images are taken on 20X magnification, how should i modify these values to work for my images.
I have attached a sample image.



#13

Hi,

What the appropriate value of the min/max diameter depends on the size of the features you want to capture. You can always estimate the diameter of objects in an image by using the distance tool; in the module display window for IdentifyPrimaryObjects (or by double-clicking the file in the File panel on the bottom-left of CellProfiler), select Tools > Measure length, then click on the image and drag to create a line. The length of the line segment is shown in the bottom-right of the display window.

If you do this for a few objects in the image for a given magnification, you can get a sense of what the min/max diameter should be.

Also, I’m attaching an adjusted version of the pipeline above, which I think should work better for the example image you posted. Basically, I manually adjusted the smoothing filter size and the maxima suppression distance, as well as changed the thresholding method used.

Regards,
-Mark
2012_11_06.cp (4.12 KB)