Hi Nuria
> Hello Gwenaëlle,
>
> I think that the mean differences between the
> protocol of your paper and the protocol described in
> the forum is that you worked with GM segmented
> data instead of pre-bet images to the halfway matrices
> (7 vs 6 DOF) application. Do you think these changes may
> increase sensibility?
The protocol described in the forum is actually exactly the protocol I've followed in my paper, I just gave more practical details. I've used each time the non-segmented images for the halfway matrices.
> With regard to Giorgio’s paper, sorry but I
> was referring to “Longitudinal changes in grey and
> white matter during adolescence” 2010, NeuroImage.
> Following it, I did the classic preprocessment VBM
> protocol with both T0 and T1, images and
> introduced the longitudinal matrix in randomise step.
> Methodologically, would you consider it correct?
I don't think there is anything really wrong with Antonio's protocol as you avoid the bias of favouring one timepoint over another one... Having said that, if your results disappear when following a more sensible and presumably more sensitive protocol, this is actually worrying. Is it just a matter of threshold of your corrected results, can you see the same pattern in the uncorrected p-values for instance?
Cheers,
Gwenaelle
--------------------------------------------------------------------
Gwenaëlle Douaud, PhD
FMRIB Centre, University of Oxford
John Radcliffe Hospital, Headington OX3 9DU Oxford UK
Tel: +44 (0) 1865 222 523 Fax: +44 (0) 1865 222 717
www.fmrib.ox.ac.uk/~douaud
--------------------------------------------------------------------
|