Hi,
Don't worry about gaps in the red lines or large patches of
red - that is all to do with the edge detector which is very simple
and sometimes misses parts of edges and also, in places like
the inter-hemispheric fissure, shows whole planes as edges
when there is a big change running across the plane.
As for areas to concentrate on, well it depends on what you
are doing (e.g. fmri, tractography, group study vs pre-surgical
planning, etc. etc.). Normally certain areas are more important
than others in particular studies and people tend to concentrate
on those so if you are interested in the caudate and the ventricles
are mis-registered then that is important, but if you are doing
a visual study you might not care so much. Note that for group
studies there is anatomical variation between individuals which
means that "exact registration" isn't really achievable and it
doesn't pay to worry too much about small mis-registrations.
Hope this helps.
All the best,
Mark
Jessica Kirkland wrote:
> Hi all,
>
> I am checking a number of registrations to the MNI template before
> segmentation, and I am wondering if anyone has advice about when
> registration may need to be re-run/tweaked. Apart from gross
> misalignment of periphery, ventricles, etc, are there more subtle red
> flags? One thing I see frequently is an incomplete red line (e.g., the
> line corresponding to the periphery may have breaks in it). Is that a
> problem? How about the blocks of red that I frequently see on a
> saggittal slice near the midline (I attempted to attach an image, but
> the file was to large to send to the list)? Alternatively, are
> there areas that are more important in terms of alignment? For
> example, if the periphery looks aligned, but cerebellum or ventricles
> seem less well-matched, etc.
>
> Thank you for the help!
>
> Jessica
>
>
|