Hi I was hoping someone could help.

When doing fMRI analysis with SPM2 I tend to do a two pass realignment
procedure in which I create a mean image and then realign to this mean
image.

I then proceed with normalising and smoothing the data.


In the first realignment when creating the mean image only I select
coregister and reslice and then mean image only. In the next realignment I
select coregister and reslice and all images (1..n), selecting the mean
image as the 1st image.

In the normalisation step I select determine parameters and write
normalised.

I was wondering if these are the options that should be selected. For
example, do I need to select the coregister and reslice options in both the
realignment steps or would coregistering be enough as they will be resliced
in the normalisation stage?

Would pushing coregister only (without reslice) make much of a difference to
the way I do it?


Furthermore as far as I know the two pass procedure minimises error, but I
think the common way to do fMRI now is to select all images + mean image in
the realignment and then use the mean image in the normalisation stage. Why
do people do it like this rather than use the two pass procedure?


My final question, if I do the two pass procedure when selecting create mean
image only and running the realignment I am presented with a file
rp_file_001.txt. The next time I run the realignment selecting all images
(1..n) a file called rp_meansfile_001.txt is created. If using movement
parameters as regressors in the analysis which text file should I use? The
one that is created when making the mean only or the one that is created
when realigning all images to the mean e.g. all images (1..n)?

Thanks in advance

Pete

_________________________________________________________________
Are you using the latest version of MSN Messenger? Download MSN Messenger
7.5 today! http://join.msn.com/messenger/overview