Hi Alessandra,
I have not looked at your results but can imagine that the output is not
satisfactory given that the procedure is suboptimal... here's a couple
of hints:
- for what you want, an initial coregistration step seems adequate, but
you should not need to reslice;
- for what you want, normalization is not likely to be the best-possible
option: it is meant to normalize one image to a template, not to match
to similar images to each other. If you really need a non-linear
matching (which very likely will distort your lesion and surrounding
tissue), you may want to look into the high-dimensional warping toolbox
instead, which will do what you want.
- if you still want to use the normalization feature, you should disable
the affine registration which likely explains some of the problems you
are experiencing. This step is tuned to matching an image to MNI space,
which involves using starting estimates which are not appropriate for
your setting. Also note that the matching is rather low-dimensional, so
some smoothing of the "target" image is likely a good idea.
- if you want to compare images of the same patient, then no
normalization is required. If you still want to normalize for some
reason, I would match the images in the subject's native space (see
above) and then normalize one of them and take along the other, using
the same set of spatial normalization parameters. This will take care of
bias due to differential spatial normalization, which has been
much-discussed on the list in the past.
There's many more things that could be said (perhaps others will chime
in), but I hope this may get you started.
Cheers,
Marko
Alessandra Carla' wrote:
> Hi,
> I have some problem with normalise.
> I have to do a non-linear coregistration between two images of the same patient.
> The original images don't have the same dimension between, so I did a linear coregistration (button coregistration estimate/write). I used one of the two images as template. I cecked the output with CheckReg, they seems to have a good alignment.
> Then I used normalise (estimate/write), the template is the same used in coregistration. The output image is very bad.
>
> The images are in http://hpmagic5.unile.it/BRAIN_IMAGE/SPM
>
> P_Normalized_biopsyBesideTumor.nii is the image used as template.
> P_Dec2011.nii is the image that I want to coregister.
> rP_Dec2011.nii is the output after coregistration.
> wrP_Dec2011.nii is the output after normalise.
>
> Could you help me? let me know where is the problem.
>
> Many thanks and best wishes.
> Alessandra Carla'
>
--
____________________________________________________
PD Dr. med. Marko Wilke
Facharzt für Kinder- und Jugendmedizin
Leiter, Experimentelle Pädiatrische Neurobildgebung
Universitäts-Kinderklinik
Abt. III (Neuropädiatrie)
Marko Wilke, MD, PhD
Pediatrician
Head, Experimental Pediatric Neuroimaging
University Children's Hospital
Dept. III (Pediatric Neurology)
Hoppe-Seyler-Str. 1
D - 72076 Tübingen, Germany
Tel. +49 7071 29-83416
Fax +49 7071 29-5473
[log in to unmask]
http://www.medizin.uni-tuebingen.de/kinder/epn/
____________________________________________________
|