Marking Cell Centroids in a Stack

[Anthony Kowal]

Hi Everyone,

I am a novice at using ImageJ, and I have tried to find this answer out but
with no success.  I have taken a time lapse recording of cells chemotaxing
towards folate, and am interested in not only finding the centroid of the
cell, but also marking it in order to help with downstream analysis.  Can
anyone please point me in the right direction on how to do this?

Thanks,

Tony

--
Tony Kowal
Graduate Student
Chisholm Laboratory

Northwestern University
303 E. Superior
Lurie Building
Room 7-250
Chicago, IL 60611

Lab: 312 503 4169
e-mail: [hidden email]

[Michael Schmid]

Hi Tony,

what do you think of, when you say "mark the centroids"? Draw a dot  
there?

If you have a stack, you could write a simple macro that runs  
"Analyze Particles" and has a loop over the particles with setPixel
(x,y) for each slice. See the nResults and getResult("Column", row)  
macro commands.

If you have binary images, e.g. created by thresholding, the "Find  
Maxima" function can create a point selection that marks the point  
closest to the centroid of each particle (in contrast to the  
centroid, this point never lies outside the particle bounds, even in  
case of a slim crescent shape). You can transfer the point selection  
to the original image by "Restore Selection" and use "Draw" to paint  
the points in the foreground color.

For getting ideas how to do this in a stack, see, e.g., the  
findStackMaxima macro
   http://rsbweb.nih.gov/ij/macros/FindStackMaxima.txt
In your case, you will have to jump between the binary image and the  
original, see the getImageID() and selectImage(id) macro functions.

I am not aware of a solution without a little bit of macro programming.

Hope this helps,

Michael
________________________________________________________________

On 12 May 2009, at 18:08, Anthony Kowal wrote:

> Hi Everyone,
>
> I am a novice at using ImageJ, and I have tried to find this answer  
> out but
> with no success.  I have taken a time lapse recording of cells  
> chemotaxing
> towards folate, and am interested in not only finding the centroid  
> of the
> cell, but also marking it in order to help with downstream  
> analysis.  Can
> anyone please point me in the right direction on how to do this?
>
> Thanks,
>
> Tony

[Anthony Kowal]

Hi Michael,

Thanks for the suggestions.  I am trying to analyze cells which are
chemotaxing towards folate, and my goal is to determine the velocity of the
cells, as well as compute a chemotactic index.  I have taken 30 to 40 minute
time lapse micrographs of these cells every 10 seconds.  For the analysis, I
am making a substack such that measurements are made every 30 seconds.  I am
using a program called MTrackJ (an Image J pluggin), but before I run
MTrackJ, I wish to mark the cell centroids (with a cross or dot), so that
when I am marking the cells I know that I am labeling the center of the
cell.

Thanks,
Tony



On 5/12/09, Michael Schmid <[hidden email]> wrote:

>
> Hi Tony,
>
> what do you think of, when you say "mark the centroids"? Draw a dot there?
>
> If you have a stack, you could write a simple macro that runs "Analyze
> Particles" and has a loop over the particles with setPixel(x,y) for each
> slice. See the nResults and getResult("Column", row) macro commands.
>
> If you have binary images, e.g. created by thresholding, the "Find Maxima"
> function can create a point selection that marks the point closest to the
> centroid of each particle (in contrast to the centroid, this point never
> lies outside the particle bounds, even in case of a slim crescent shape).
> You can transfer the point selection to the original image by "Restore
> Selection" and use "Draw" to paint the points in the foreground color.
>
> For getting ideas how to do this in a stack, see, e.g., the findStackMaxima
> macro
>  http://rsbweb.nih.gov/ij/macros/FindStackMaxima.txt
> In your case, you will have to jump between the binary image and the
> original, see the getImageID() and selectImage(id) macro functions.
>
> I am not aware of a solution without a little bit of macro programming.
>
> Hope this helps,
>
> Michael
> ________________________________________________________________
>
> On 12 May 2009, at 18:08, Anthony Kowal wrote:
>
>  Hi Everyone,
>>
>> I am a novice at using ImageJ, and I have tried to find this answer out
>> but
>> with no success.  I have taken a time lapse recording of cells chemotaxing
>> towards folate, and am interested in not only finding the centroid of the
>> cell, but also marking it in order to help with downstream analysis.  Can
>> anyone please point me in the right direction on how to do this?
>>
>> Thanks,
>>
>> Tony
>>
>

--
Tony Kowal
Graduate Student
Chisholm Laboratory

Northwestern University
303 E. Superior
Lurie Building
Room 7-250
Chicago, IL 60611

Lab: 312 503 4169
e-mail: [hidden email]