Hi,
Do you have much neck in the images or are they in
non-axial orientations? These are both common
causes of registration problems, most of which can be
solved by initially reslicing the image to axial and/or
removing the neck. You can reslice to axial using:
fslswapdim orig RL PA IS new
turns the image "orig" into "new", which should be
in axial orientation (like the MNI152).
Once you've done this (if necessary) you can then
look at the image in FSLView, find out where the
brain starts and ends (in the z axis) and use fslroi
to remove the neck.
This is probably the easiest way to solve the problems
if either of these conditions pertains to your images.
If not please let us know and we'll try to sort it out.
All the best,
Mark
On 10 Sep 2009, at 19:55, Maxim Bester wrote:
> Dear FSLers
>
> During a recent study using SIENAX i frequently ( almost every second
> subject) run into the problem that the images (T1 MPRAGE at 3T) show
> very
> bad coregistration to MNI. is there any possibility to pass commands
> to
> flirt in the sienax command options (or even any other known way to
> solve
> this problem) ?
>
> thanks
>
> maxim
>
|