Dear Daniel,
As Saad wrote the other day:
> In general, combining different protocols is not recommended. But if your aimed contrasts are orthogonal to protocol type, then you may be able to combine the data. (for instance for a group difference where the groups are distributed equally in both protocols).
> Your TR difference is unlikely to cause major differences, as T1 recovery should have had sufficient time to take place.
> You don’t need to resample the data, the TBSS pipeline will do it for you.
I would recommend using a covariate that indicates the acquisition group in order to remove any bias.
All the best,
Mark
On 9 Jan 2014, at 08:01, Daniel Ferreira <[log in to unmask]> wrote:
> Dear experts,
>
> I have a DTI dataset of 73 subjects with some variability in the sequence parameters:
>
> Repetition time: 19 subjects were given a TR of 15000 while the rest got a TR of 17000.
>
> Echo time: most of subjects got a TE around 72, but there are some ranging from 73 to 87. Even there is one with a TE of 96 and another one with 107.
>
> I wonder if it is possible to combine them for TBSS analyses. In case it is, should I enter TR subgroup as covariate or is it not necessary? What about the TE?
>
> Thanks so much and happy 2014!
> _______________
> Daniel Ferreira, PhD.
> Department of Neurobiology, Care Sciences and Society, Division of Clinical Geriatrics (Karolinska Institute, Sweden)
> [log in to unmask] | [log in to unmask]
|