Matthew -
It's a good question, and I'm still not sure what the correct answer is.
If the temporal interpolation were perfect, it would not matter which
reference slice you interpolated the data to (though it would of course
matter that you synchronised the model with that reference slice).
However, the full fourier interpolation is not perfect - it will alias
frequencies above the Nyquist limit. In this case, I would imagine
(though am not certain) that the interpolation error will increase the
more you try to shift the data, ie for slices further from the reference
slice (it certainly should have no effect for the reference slice
itself, which is unchanged by the interpolation). If so, the middle
slice
would be the best default.
Others have argued that negligible power exists in the haemodynamic
response
above typical Nyquist limits, and therefore the reference slice doesn't
matter,
but more generally this question depends on 1) your TR and 2) the
frequency
structure of your experimental design.
The issue was raised in the thread:
http://www.mailbase.ac.uk/lists/spm/1999-06/0005.html
http://www.mailbase.ac.uk/lists/spm/1999-06/0018.html
Best wishes
Rik
Matthew Brett wrote:
>
> Dear Rik, Jesper and all,
>
> Ah, sorry, rereading my message, it wasn't atall clear what I meant. We
> routinely use the slice timing routine to resample to the first slice in
> the volume. I can see that this may differ from a correction to the
> central slice of the volume, but wasn't sure whether this made any
> difference in practice.
>
> Thanks a lot,
>
> Matthew
Rik Henson wrote:
>
> > Can I ask though; do you have a feel for the importance of centering the slice
> > timing on the middle slice rather than the first? Has anyone assessed its
> > impact on real T maps?
>
> For blocked designs, the longer the blocks, the less important
> the slice-timing. For event-related designs, it can make a big
> difference (with TR>2s) - see the Henson et al abstract that
> Jesper cited.
>
> Rik
>
> ---------------------------8-{)}-------------------------
>
> DR R HENSON EMAIL [log in to unmask]
> Wellcome Department of
> Cognitive Neurology TEL (work1) +44 207 833 7483
> 12 Queen Square TEL (work2) +44 207 833 7472
> London, WC1N 3BG FAX +44 207 813 1420
>
> URL: http://www.psychol.ucl.ac.uk/rik.henson/index.html
>
> ---------------------------------------------------------
--
---------------------------8-{)}-------------------------
DR R HENSON EMAIL [log in to unmask]
Wellcome Department of
Cognitive Neurology TEL (work1) +44 207 833 7483
12 Queen Square TEL (work2) +44 207 833 7472
London, WC1N 3BG FAX +44 207 813 1420
URL: http://www.psychol.ucl.ac.uk/rik.henson/index.html
---------------------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|