Intensity quantification area

[Rbalada]

Hello everybody,

I'm a first year PhD student from Barcelona and I'm having some doubts
using ImageJ. If anybody could help answering my questions I'd be really
grateful. Thanks a lot and enjoy your research.

So, I'm trying to quantify the intensity of an Immuno Histo Quemistry from rat hippocampus and I'd like to measure it using always the same rectangle, with the same exact area, so then I can compare the different intensities regarded in each picture from each condition. That's my first doubt, how do I get sure of using always the same rectangle?

Second question: how can I correlate each intensity point within a rectangle with the exact point of the picture? (the coordenates). For what I've been told, ImageJ homogenizes the different intensity points within a rectangle, so you can not know something like 'at the coordenate x= 0,5 and y= 1,3 the intensity is...whatever. So, is that like this or there is an option to see each intensity value within a rectangle? If so, how could I do this? Any idea of how could I know this intensity values using some other program if ImageJ does always homogenize the intensity within a rectangle? Maybe the homogenized intensity within a rectangle may work for comparing between conditions, but not between different points within the same rectangle of the same condition.

Thanks again and please, if my way of explaining my situation with ImageJ is not clear but you think you may be able to help me if I explained it better, don't doubt asking me to explain it again, I'll be really grateful to you.

Thanks a lot,




Rafel

[Gabriel Landini]

On Thursday 31 Oct 2013 10:40:29 Rbalada wrote:

> So, I'm trying to quantify the intensity of an Immuno Histo Quemistry from
> rat hippocampus and I'd like to measure it using always the same rectangle,
> with the same exact area, so then I can compare the different intensities
> regarded in each picture from each condition. That's my first doubt, how do
> I get sure of using always the same rectangle?
>
> Second question: how can I correlate each intensity point within a rectangle
> with the exact point of the picture? (the coordenates). For what I've been
> told, ImageJ homogenizes the different intensity points within a rectangle,
> so you can not know something like 'at the coordenate x= 0,5 and y= 1,3 the
> intensity is...whatever. So, is that like this or there is an option to see
> each intensity value within a rectangle? If so, how could I do this? Any
> idea of how could I know this intensity values using some other program if
> ImageJ does always homogenize the intensity within a rectangle? Maybe the
> homogenized intensity within a rectangle may work for comparing between
> conditions, but not between different points within the same rectangle of
> the same condition.

Unfortunately it is not possible to quantify immunohistochemistry reliably.
See the "Note" in red here:
http://www.dentistry.bham.ac.uk/landinig/software/cdeconv/cdeconv.html

You might be able to capture some intensity of stain, but that would not tell
reliably much about how much product there is present which relates to the
amount of antigen present.

To be able to quantify the amount of products, you need a stoichiometric
reaction (IHC is not so) AND a chromogen that follows Beers-Lambert law.

Cheers

Gabriel

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

[Heeschen, Bill (WA)]

Greetings, Rafel.
[I use digest mode, so hopefully this note is not 100% redundant!]
All the raw capabilities for which you ask are available in ImageJ.  However, Gabriel's comment about quantitation MUST be heeded: You can, in fact, get the numbers from the images, but you must be careful intepreting them.

The "rectangle" you are using is known as a region-of-interest or ROI.  If you look through the ImageJ documentation for ROI handling you will find that they are extremely useful and flexible.  The tools for dealing with ROIs are quite powerful.
1)      Setting a rectangle to the exact same location in subsequent images can be done two ways:
a.      Draw a rectangle (or any other type of ROI) in one image.  Open (or switch to) another image and select the Edit=>Selection=>Restore Selection submenu option.  Note that pressing the CTRL-shift-E key combination does the same action.  The selection from the first image will be exactly reproduced in the second.
b.      Use the ROIManager:  Draw a rectangle in an image.  Select the Edit=>Selection=>Add to Manager submenu option.  The selection will now be stored in the ROIManager window.  If the ROIManager is not already open, it will open when you add the selection. Clicking on the ROI label in the ROIManager will cause that ROI to be reproduced in the current (top-most) image.  You can store the selections in the ROIManager for later use by clicking the More... button, then selecting the Save... option. Use the normal ImageJ Open... command to open a saved ROI set.
2)      Getting pixel-by-pixel data is easy to do manually.  Getting an automatic report for all the pixels within an ROI requires a macro, but the macro would be pretty straightforward to write.  For manual extraction there are two ways:
a.      Use your mouse (no clicking) and hover over the pixel you want to measure. Watch the console message area at the bottom of the ImageJ control panel (below the tool icons).  As you hover over pixels, the location in calibrated units shows up as the X=  and Y=  values.  The intensity of the pixel is also included.  If it is a gray-scale image, only the gray value shows up.  If it is a color image, the values for the red, green and blue planes are listed.  If the intensity is calibrated, the value also shows up in calibrated units (like optical density).
b.      Use the Point selection tool (plus sign with a square in the middle). Double-click on the tool to select Auto-Measure and whatever other options you want.  Move over the pixel you want to measure and click on it.  The data about that pixel will show up in your Results window.

So, my suggestion is that you read the ImageJ documentation about ROIs and measurement.  You will discover that ImageJ by itself is capable of a very wide range of measurements.  If you look over the plugins and macros that have been contributed by the user community, you will be thoroughly overwhelmed by what the software can do.  The really great thing is that it is open source, so all it takes is some programming ability to make it do anything!

ImageJ website: http://imagej.nih.gov/ij/
Documentation page: http://imagej.nih.gov/ij/docs/index.html

Best Regards,

Bill
William A. Heeschen, Ph.D.
Microscopy, Digital Imaging
The Dow Chemical Company
Midland, MI  48667
[hidden email]



Date:    Thu, 31 Oct 2013 03:40:29 -0700
From:    Rbalada <[hidden email]<mailto:[hidden email]>>
Subject: Intensity quantification area

Hello everybody,

I'm a first year PhD student from Barcelona and I'm having some doubts
using ImageJ. If anybody could help answering my questions I'd be really
grateful. Thanks a lot and enjoy your research.

So, I'm trying to quantify the intensity of an Immuno Histo Quemistry from
rat hippocampus and I'd like to measure it using always the same rectangle,
with the same exact area, so then I can compare the different intensities
regarded in each picture from each condition. That's my first doubt, how do
I get sure of using always the same rectangle?

Second question: how can I correlate each intensity point within a rectangle
with the exact point of the picture? (the coordenates). For what I've been
told, ImageJ homogenizes the different intensity points within a rectangle,
so you can not know something like 'at the coordenate x= 0,5 and y= 1,3 the
intensity is...whatever. So, is that like this or there is an option to see
each intensity value within a rectangle? If so, how could I do this? Any
idea of how could I know this intensity values using some other program if
ImageJ does always homogenize the intensity within a rectangle? Maybe the
homogenized intensity within a rectangle may work for comparing between
conditions, but not between different points within the same rectangle of
the same condition.

Thanks again and please, if my way of explaining my situation with ImageJ is
not clear but you think you may be able to help me if I explained it better,
don't doubt asking me to explain it again, I'll be really grateful to you.

Thanks a lot,




Rafel



--
View this message in context: http://imagej.1557.x6.nabble.com/Intensity-quantification-area-tp5005401.html
Sent from the ImageJ mailing list archive at Nabble.com.

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






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

[Rbalada]

Hi Bill,


Thanks a lot for your help and information, I really appreciate it. I'll read the documents and I'll follow the method you suggest, hopefully it will work for what I need.


Cheers,



Rafel

[Rbalada]

Hello Gabriel,


Thanks a lot for your help. I need something to compare the different conditions between them, so even if I can not quantify IHQ reliably, maybe I can get some data that I can use to compare, not using the 'real' amount of product but using the 'intensity ImageJ reads' or, as you say, the intensity of stain in each condition, which in addition of other techniques may give more strengh to the results (that's something I'll have to discuss later and your answer will be important for me to explain what can be used for what).

So, thanks a lot for your words and information, it made me realize about what I was dealing with (or maybe more exactly, what I was not).


Cheers,



Rafel