effect of magnification on area measurements

[Cheuk Tam]

Dear ImageJ Community,

I am new to ImageJ and am using it to measure the area of DAB-stained
cells.  I tried measuring the area of the same cell at different
magnifications (and set the scale to the respective stage micrometer
images), but I found the area measurements yielded different numbers.  The
40X objective had an area that was about 4-fold smaller than the 10X
objective image.  Is this because the resolution and the number of pixels in
each image are different?  If so, which area measurement is the accurate
one?   Thank you for your help.

-Cheuk

[Nathaniel Ryckman]

"and set the scale to the respective stage micrometer images"

What do you mean by that statement? Just to check, do you mean that you changed the scale for each image that you were looking at by using Analyze->Set Scale?

Could you provide us with the settings you used for each image?

Cheuk Tam wrote

Dear ImageJ Community,

I am new to ImageJ and am using it to measure the area of DAB-stained
cells.  I tried measuring the area of the same cell at different
magnifications (and set the scale to the respective stage micrometer
images), but I found the area measurements yielded different numbers.  The
40X objective had an area that was about 4-fold smaller than the 10X
objective image.  Is this because the resolution and the number of pixels in
each image are different?  If so, which area measurement is the accurate
one?   Thank you for your help.

-Cheuk

[Nathaniel Ryckman]

My guess is that you used the same scale for both the 40x and 10x image?

You'll have to use a different scale for each of those images. One scale will be 4x greater than the other scale.

Basically your picture is broken up into a grid of pixels (dots). You have to specify the length that each dot represents. In the 40x image, the grid of pixels hasn't changed (aka your resolution), but the amount of distance that each dot represents has changed.

Cheuk Tam wrote

Dear ImageJ Community,

I am new to ImageJ and am using it to measure the area of DAB-stained
cells.  I tried measuring the area of the same cell at different
magnifications (and set the scale to the respective stage micrometer
images), but I found the area measurements yielded different numbers.  The
40X objective had an area that was about 4-fold smaller than the 10X
objective image.  Is this because the resolution and the number of pixels in
each image are different?  If so, which area measurement is the accurate
one?   Thank you for your help.

-Cheuk

[pang]

You need to use Analyze/Set Scale function of the ImageJ. As long as you
set the scale correctly, you should get the same area for two images at
different magnification. ImageJ itself does not know the scale unless
you set it.

I suspected that you use the Image/Scale function, this will
reduce/enlarge your image.


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



-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
Cheuk Tam
Sent: Friday, April 15, 2011 1:54 PM
To: [hidden email]
Subject: effect of magnification on area measurements

Dear ImageJ Community,

I am new to ImageJ and am using it to measure the area of DAB-stained
cells.  I tried measuring the area of the same cell at different
magnifications (and set the scale to the respective stage micrometer
images), but I found the area measurements yielded different numbers.
The 40X objective had an area that was about 4-fold smaller than the 10X
objective image.  Is this because the resolution and the number of
pixels in each image are different?  If so, which area measurement is
the accurate
one?   Thank you for your help.

-Cheuk

[Gabriel Landini]

On Friday 15 Apr 2011, Cheuk Tam <[hidden email]> wrote:
> I am new to ImageJ and am using it to measure the area of DAB-stained
> cells.  I tried measuring the area of the same cell at different
> magnifications (and set the scale to the respective stage micrometer
> images), but I found the area measurements yielded different numbers.  The
> 40X objective had an area that was about 4-fold smaller than the 10X
> objective image.  Is this because the resolution and the number of pixels
> in each image are different?

Are you sure that your calibration is correct? I would look at this first.
Second I would measure a single object that is easily defined and has s simple
shape - let's say 1 nucleus  or the same erythrocyte (do you get the right
diameter at all magnifications?).

If you still got the wrong numbers, you probably have done wrong the
calibration of the objectives or made a mistake with the photos thinking that
you used a different objective. Or you did not apply the right calibration to
the image. Or,  the thresholds you apply to the different images are selecting
different things.

It also matters what you are measuring and what shape and how close your
objects are.
Euclidean shapes should return a similar area size across maginifications plus
minus some error associated to the digitisation and inter-pixel distance.
The numbers in the stage micrometer are a good test because they are well
contrasted.
If what you are measuring is not Euclidean, maybe at low magnification many
objects merge into one, or empty holes appear filled at low magnification.
Without any other details or images (please upload them to some site for us to
see) it is not possible say much more.

Cheers

G