I think the main thing is to distinguish coregistration (which is
simple) and spatial normalization (which is somewhat more complex). Some
> I am analyzing the fMRI data for different subjects individually(
> single session only )
If you want to do single subjects only, why normalize in the first
place? You only need to normalize if you want to compare/summarize
activations across individuals.
> 1. I want to map my data to some standard Template(I chose MNI-ICBM
> 152 Linear) and for that I'm using Coregister(ESTIMATE) method
> firstly with mean functional image(as reference) and MNI template(as
> source) and secondly with mean functional image(as reference) and
> structural image(as source). Am I doing it correctly?
No: coregistration is within-subject, between-modality registration. You
can coregister a subject's T1 to his/her T2. Some people, myself
included, also use coregistration of an individual's T1 to a template in
order to broadly align it along the main axes, but not in order to
> here because as far as I think changing the orientation of MNI with
> respect to mean image is not similar as vice-verse. And I am not
> keeping the mean image as source because that will change the
> orientation of mean image with respect to other images. If I am
> correct here then how should I proceed?
As mentioned above, you do not want to coregister here anyway. But if
you, for example, use coregistration for the mean EPI to the anatomical,
you should specify all functionals as "other images" in the batch file.
They are then "taken along" if you move the mean EPI to match the T1.
> 2. For normalisation I am using the parameter file generated by
> Segment method. From what I read in the manual Normalise method also
> orients data into some standard space defined by templates. In my
> case is this similar to what coregister will do with the data? If
> yes, then if instead of using parameter file generated by segment
> method, I use the above mentioned standard template(MNI-152) for
> normalization, will it affect my analysis?
Whichever choice you select in processing will affect your analysis,
sometimes in more, sometimes in less obvious ways. In your case, you
should realize that unified segmentation (which is what happens if you
click "segment" in spm8) will also normalize the data, but based on the
prior tissue probability maps specified in the batch. These may, or may
not, match the MNI152 template. You can then use the generated
parameters to normalize the functionals (or only the con-images of the
native space analyses), but again, only if you want to compare
activations in the first place.
PD Dr. med. Marko Wilke
Facharzt für Kinder- und Jugendmedizin
Leiter, Experimentelle Pädiatrische Neurobildgebung
Abt. III (Neuropädiatrie)
Marko Wilke, MD, PhD
Head, Experimental Pediatric Neuroimaging
University Children's Hospital
Dept. III (Pediatric Neurology)
D - 72076 Tübingen, Germany
Tel. +49 7071 29-83416
Fax +49 7071 29-5473
[log in to unmask]