Dear Martin,
The refvol is used to determine the ROI used for calculating the
rms difference over. Essentially it specifies the centre for a
sphere of radius 80mm. It normally doesn't make a huge
difference to the results what is chosen, unless there are big
rotations and translations. I often use the input image here,
especially if there is a huge difference in FOV, but as I said
it doesn't matter much.
If you want to measure the "amount" of transformation going
on then comparing with the identity matrix is very sensible.
All the best,
Mark
On 22 Oct 2011, at 04:25, Martin M Monti wrote:
> Dear FSLers,
>
> I have two quick questions on the use of rmsdiff: I am trying to get a single measure of "how much" sheering/translating/etc there was across a set of (linear) registrations.
> 1. I am a little confused by what would be 'refvol' in the command's syntax. In other commands refvol is the "to" image(i.e., in example_func2highres it would be the "highres"), but on the manual online it's referred to as an "input image" which would make me think of "example_func". So is it the "from" or "to" image I need there?
> 2. I am not comparing different registrations for the same image, rather I would just like to get a number as to, for each of my subjects, how much moving around there was. Would it be fair game to use an identity matrix (for all my subjects) as the second input matrix? Naively, I'd think this would give me a common "no registration" benchmark for all the subjects.
>
> thank you
>
> Martin
>
>
> --
> Martin M Monti, PhD
> Assistant Professor
> 7461E Franz Hall
> UCLA Department of Psychology
> BOX 951563,
> Los Angeles, CA 90095-1563
> 310-825-8546
> http://montilab.psych.ucla.edu
> http://faculty.psych.ucla.edu/directory/faculty.php?id=187&area=3 <http://faculty.psych.ucla.edu/directory/faculty.php?id=187&area=3%20>
>
> “Well! I’ve often seen a cat without a grin,” thought Alice,
> “but a grin without a cat! It’s the most curious thing I ever
> saw in my life!” [Lewis Carroll, Alice in Wonderland]
>
|