The option is there because I thought it would be useful to a few
people. I don't have so much experience of actually using it, but
suspect that you should be able to realign and reslice the data in
such a way that the nose is moved back to the right place. This may
involve making a copy of one of your scans, which you'd then translate
in the phase encode direction. The remainder of the scans could then
be resliced such that they are brought into alignment with it.
Spatial normalisation should probably not be done with wrapping
enabled. The reason for this is that you may see the image repeated
in parts of the normalised data.
On 13 July 2011 13:08, Ramsey Raafat <[log in to unmask]> wrote:
> Dear John (and SPMers),
> We have the nose "wrapped", sitting behind at the back of the head and
> hence also perhaps effecting the EPI’s, as the data is acquired
> periodically I believe this is not a problem and just a question of
> wrapping the data around on itself. I have enquired with the local
> expertise, however I am unsure whether this wrap has to be applied at the
> realign stage, that is do you consistently have to enable Wrap in the
> defaults( or via a script) eg say defaults.realign.write.wrap [ 0 0 0] ,
> and in all subsequent pre-processing operations, that is, does it "carry
> through", or do all the default wrap operations have to be enabled ( eg
> defaults.coreg.write.wrap , defaults.normalise.write.wrap etc). From my
> understanding of the code, the zero vector above [ 0 0 0] refers to the
> directions x,y z, so if the nose is wrapped around in the phase encode
> direction or the y direction would that indicate that the vector should
> be [ 0 1 0] to enable wrapping?( and finally if time permits are there any
> adverse effects to having the wrap around enabled as a default, or should
> this be avoided?)
> Many thanks in advance!
> With very best wishes!