Fluorescent measurements

[Madaan, Uday]

Dear All,

I am using imagej to measure the fluorescence of immunofluorescent staining.
I am currently using the RGB Color plugin under graphics.
My question is:
1) Is this the correct way to measure fluorscence intensity using ImageJ. Is there a better way to do this?
2) When I measure I get a RGB profile graph. On that graph I get a line graph which peaks and comes down.
How exactly is the fluorscent intesity being recorded? Why does the line graph peak and descend?

Thank you for all your help.

Best Regards,
Uday
Uday Madaan
[hidden email]

_______________________________________________________

ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[pang]

Uday,

Fluorescence images are usually pseudo -colored, and it is better to use
the original monochrome image to do the measurement.  Most camera for
fluorescence microscope is monochrome camera because it is more
sensitive than color camera.

If you already got  a color image, and you use R for one staining, G for
another stain, and B for the last stain (or no stain). You can separate
RGB channel into individual channels and then quantify them.

It is better to start with monochrome camera and process the raw image.
If needed, you can send me the image and we can look together.

Zhengyu Pang

Zhengyu Pang, Ph.D.
Biochemistry and Bioanalytics Laboratory
Diagnostic and Biomedical Technologies
GE Global Research, K1-5B37A
One Research Circle
Niskayuna, NY 12309
T: 518-387-4015
F: 518-387-7765



-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
Madaan, Uday
Sent: Thursday, May 31, 2012 3:51 PM
To: [hidden email]
Subject: Fluorescent measurements

Dear All,

I am using imagej to measure the fluorescence of immunofluorescent
staining.
I am currently using the RGB Color plugin under graphics.
My question is:
1) Is this the correct way to measure fluorscence intensity using
ImageJ. Is there a better way to do this?
2) When I measure I get a RGB profile graph. On that graph I get a line
graph which peaks and comes down.
How exactly is the fluorscent intesity being recorded? Why does the line
graph peak and descend?

Thank you for all your help.

Best Regards,
Uday
Uday Madaan
[hidden email]

_______________________________________________________

ImageJ mailing list: http://imagej.nih.gov/ij/list.html

_______________________________________________________

ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Cammer, Michael]

The remove outliers command is great, especially when our images are spatially way oversampled, but I was wondering whether it would be possible to modify it with a hybrid for replacing the pixel.  Right now the pixels are made black when removing high value outliers.  Is there a way we could have a hybrid to replace the center pixel with the average or median of its neighbors?  I could write a macro to do this in a second pass (add 1 to the image and then look for pixels of value 0 to operate on), but it would be much faster to have this included in the command to begin with.
Thanks!!

________________________________________________________
Michael Cammer, Assistant Research Scientist
Skirball Institute of Biomolecular Medicine
Lab: [phone]  Cell: [phone]

_______________________________________________________

ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Stephan Saalfeld]

Hi Michael,

have a look at the Fiji plugin

Plugins > Integral Image Filters > Remove Outliers

it uses a different approach and smears border pixels into the holes.
For small holes that can look very convincing.  The plugin is also very
very quick, you should check the preview box to test parameters.  In
this plugin, the threshold is specified as a multiple of the
standard-deviation.  The standard-deviation is calculated for each pixel
independently using a block of given radius around it.  Thanks to
integral images the speed is invariant to the block size.

Best,
Stephan




On Fri, 2012-06-01 at 18:40 +0000, Cammer, Michael wrote:

> The remove outliers command is great, especially when our images are spatially way oversampled, but I was wondering whether it would be possible to modify it with a hybrid for replacing the pixel.  Right now the pixels are made black when removing high value outliers.  Is there a way we could have a hybrid to replace the center pixel with the average or median of its neighbors?  I could write a macro to do this in a second pass (add 1 to the image and then look for pixels of value 0 to operate on), but it would be much faster to have this included in the command to begin with.
> Thanks!!
>
> ________________________________________________________
> Michael Cammer, Assistant Research Scientist
> Skirball Institute of Biomolecular Medicine
> Lab: [phone]  Cell: [phone]
>
> _______________________________________________________
>
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

_______________________________________________________

ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Michael Schmid]

Hi Michael,

Process>Noise>Remove Outliers should replace the pixel values of the
outliers with the median within the 'radius' given in the dialog.  If it
does not, this sounds like a bug; could you mail a sample image to me
off-list? (together with the radius&threshold values you are using).

Michael
___________________________________________________________________

On Fri, June 1, 2012 20:40, Cammer, Michael wrote:

> The remove outliers command is great, especially when our images are
> spatially way oversampled, but I was wondering whether it would be
> possible to modify it with a hybrid for replacing the pixel.  Right now
> the pixels are made black when removing high value outliers.  Is there a
> way we could have a hybrid to replace the center pixel with the average or
> median of its neighbors?  I could write a macro to do this in a second
> pass (add 1 to the image and then look for pixels of value 0 to operate
> on), but it would be much faster to have this included in the command to
> begin with.
> Thanks!!
>
> ________________________________________________________
> Michael Cammer, Assistant Research Scientist
> Skirball Institute of Biomolecular Medicine
> Lab: [phone]  Cell: [phone]
>
> _______________________________________________________
>
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

_______________________________________________________

ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Stephan Saalfeld]

Hi Bill,

The code is in the mpicbg library that fiji includes as a submodule

https://github.com/axtimwalde/mpicbg

I put a description of the methods on the Fiji Wiki:

http://fiji.sc/wiki/index.php/Integral_Image_Filters

The page has a link to where the actual code lives.  Check the class
BlockStatistics for local mean, std, variance and sample variance.  The
Outlier Removal Plugin together with friends is here

http://fiji.sc/cgi-bin/gitweb.cgi?p=mpicbg.git;a=tree;f=mpicbg/ij/plugin

making the ImageJ entry points into the actual implementations.

Best,
Stephan




On Sun, 2012-06-03 at 11:41 -0400, Bill Christens-Barry wrote:

> Stephan,
>
> I'm interested in looking at the source for the plugin you mentioned.
> As a non-expert java coder, I'm a little bewildered with how things
> are organized in a project like this, and my read of docs at the fiji
> site hasn't gotten me far yet. After cloning the git fiji project, how
> can I find the relevant source? I don't see any files with names that
> I would naively expect to be used. I'm particularly interested in the
> code used for the integral imaging methods and in the computation of
> local variance/standard deviation.
>
> Thanks.
>
> Bill Christens-Barry
> [hidden email]
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Cammer, Michael]

Dear all,
Apologies for raising this false alarm.  I should have been more careful in my analyses before posting to the group.  The function works fine.
Regards,
Michael

-----Original Message-----

On Jun 1, 2012, at 2:40 PM, Cammer, Michael wrote:

> The remove outliers command is great, especially when our images are spatially way oversampled, but I was wondering whether it would be possible to modify it with a hybrid for replacing the pixel.  Right now the pixels are made black when removing high value outliers.  Is there a way we could have a hybrid to replace the center pixel with the average or median of its neighbors?  I could write a macro to do this in a second pass (add 1 to the image and then look for pixels of value 0 to operate on), but it would be much faster to have this included in the command to begin with.
> Thanks!!
> ________________________________________________________
> Michael Cammer, Assistant Research Scientist Skirball Institute of
> Biomolecular Medicine
> Lab: [phone]  Cell: [phone]
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html