hi Wei
> I raised this question a few days ago but didn't get an answer. I assume
> it's a very simple question or maybe I'm doing something wrong. I analyzed
> the motor task data on a single subject. I normalized the coregistered
> anatomical (to EPIs) to the MNI standard template. Now I want to display
> the activation result on top of the subject's anatomical, instead of the
> MNI. How do I extract the normalization information to get the activation
> map aligned in the subject's anatomical space? Or is there a different
> process I should process the data to keep the activation map in the same
> space of its own anatomical?
There are a few ways to do this, all of which I suspect will give you
a fairly similar result:
1) Normalize the subject's structural image (using the same parameters
as you used for the functional images) to MNI space and display the
statistics on that;
2) If you've normalized the images using the unified segmentation
approach (i.e. the "segment" button in SPM5 or SPM8), you should have
both a *seg_sn.mat (giving the transformation from subject space to
MNI space) and *seg_inv_sn.mat, which gives the opposite
transformation. If you apply the *seg_inv_sn.mat parameters to the
MNI-space images, they should be in subject space, the same as the
anatomical;
3) It would require re-running your first level models, but you can do
all your statistics on the un-normalized (but realigned and smoothed)
EPI images, so everything will be in subject space. If you want to do
group statistics on these you can just normalize the con* images from
your first level contrasts, and put these in a group analysis.
There are probably also a few older emails floating about on what you
want to do; see for example
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0808&L=SPM&P=R9881
which is approach (2) above.
Hope this helps,
Jonathan
|