Hi
The two are essentially equivalent given that DTIFIT does a sum of squares minimisation with a GLM.
(I say essentially because in DTIFIT, the log(data) is used for the fitting, so you won’t get exactly the same result when you concatenate versus average, but they’ll be pretty close)
Cheers,
Saad
> On 20 Jul 2017, at 17:17, Lucas Lessa <[log in to unmask]> wrote:
>
> Dear colleagues,
>
> I did two times the DTI scanning in my reasearch, to increase the signal/ratio of the data.
> Now I want to analyse by TBSS the differences between the groups, but using all the files.
> I thought about using FSLMATHS to average the files, after eddy correction, and in merging all the files per subject into one and replicate the bvals and bvecs.
> Wich one do you think it's best?
> Fell free to sugest another to do this anlyses also.
> Thanks.
>
> Lucas Lessa.
|