John, Christian, and others with experience in this:
One question that comes to mind when considering a masking procedure
as suggested here:
https://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind0908&L=SPM&P=R895
is that the values that get masked out (i.e. the streaky bit at the
top) came from somewhere; wouldn't this suggest that the data left
where we want it (in the brain!) is going to be affected? At the very
least, affected differently across subjects, especially where the
exact amount of partial volume coverage might differ due to varying
head size?
My naive assumption would be that regions nearer to the streaky
out-of-the-brain region would be more affected, since they would be
more likely to contribute to the data that end up getting masked.
Jonathan
On Thu, Oct 1, 2009 at 6:57 PM, John Ashburner <[log in to unmask]> wrote:
> It looks entirely as expected to me. The streaky bit at the top is the
> code trying to fill in some unknown values, as this part of the image is
> missing in the original fMRI data.
>
> See the various other posts on the mailing list (eg by Christian
> Buchel), which recommend generating a suitable mask for the statistical
> analysis.
>
> Best regards,
> -John
>
|