choosing the right ROI

Posted by Luciano Marpegan on Mar 06, 2008; 12:57am
URL: http://imagej.273.s1.nabble.com/save-as-jpg-tp3696972p3696987.html

Hi everybody,
I have to measure the signal intensity from cells in time-lapse
movies. The cells move and their intensity and shape also change from
frame to frame. The shape is mostly circular so we use circular ROIs
to perform the measurements but if the diameter changes then we face
the problem of changing the size of our ROI or not. The way we solve
this is by using the same ROI for all the cells (we choose it so that
the biggest cell will fit in it) through all the movie, meaning that
for most of the cells we are also measuring a lot of the background.
The parameter we measure is integrated intensity.
Is there a better way to approach this type of measurements? I've
seen people using the 10 brightest pixels within the ROI or getting
the histogram within the ROIs and comparing their distributions but I
don't have the knowledge to determine what "should" be done and which
are the pros and cons of the different techniques. If there is
literature on this topic I would appreciate any suggestions.

Many Thanks,
Luciano