On Sat, 20 Oct 2007 01:01:48 -0400, Buyean Lee <[log in to unmask]> wrote:
>
>Dear SPM users,
>
>I have checked the email list to find the answer for my simple question,
but I couldn't.
>
>Somehow, PET to MRI coregistration is better than MRI to PET image, at
least in my animal images.
>
>
>
> Would someone let me know how to apply the inverse transformation?
The basic idea should be to take the matrix in the *.mat file and take the
matrix inverse of it.
I say "basic," because I'm leaving out some details. (E.g.: in SPM2,
there are two matrix variables, "M" and "mat" if I recall correctly. Then
there's the issue of if anything additional need to be done if the two
images have difference mesh sizes, origins, etc, which I haven't thought
about.)
>
>In other words, I would like to generate the transformation matrix by
running PET (source image) to MRI (reference image) coregistration, but I
want to coregister MRI to PET by using the transformation matrix generated
by PET to MRI coregistration like the general advice found in the FSL
website (see below).
>
>I would like to do this in both SPM2 and SPM5.
>
>"Use the image with the best quality (tissue contrast and resolution)
>as the Reference image, otherwise poor registrations may occur (the
inverse
>transformation can be found afterwards using the InvertXFM gui or the
convert_xfm command-line utility)"
>
>Thank you,
>
>Buyean
>
>________________________________________________________________________
>Check Out the new free AIM(R) Mail -- Unlimited storage and industry-
leading spam and email virus protection.
>
|