>> Are there any pathologies out there that causes sprouting of new
>> brain tissue?
> Yes, abstinence;)
That's very republican of you ;-)
Puss J
>
> Cheers-
> Andreas
>
>
> ________________________________________
> Von: FSL - FMRIB's Software Library [[log in to unmask]] im Auftrag
> von Jesper Andersson [[log in to unmask]]
> Gesendet: Mittwoch, 27. Mai 2009 13:36
> An: [log in to unmask]
> Betreff: Re: [FSL] Sienar Vs FSL-VBM
>
> Dear Antonios,
>
>> I have some questions regarding the similarities and differences
>> between
>> SIENAr and FSL-VBM.
>>
>> Correct me if I'm wrong , SIENAr will measure group differences only
>> in gray
>> matter areas which are located in brain edges or very close to brain
>> edges
>> (since flow images are dilated) while from the other side FSL-VBM is
>> able to
>> measure differences in all gray matter voxels defined by the gray
>> matter
>> mask (produced in the first stages of VBM's analysis). So a direct
>> comparison of voxelwise changes -with both methods in a specific
>> data set
>> for gray matter differences- will not present any overlapping areas
>> (except
>> only from the brain edges). Is my assumption correct?
>
> SIENAr will only detect changes on/very near the cortical surfaces and
> at the CSF-white matter junctions of the ventricles. It is
> specifically designed to look for atrophy, so that is hardly a
> limitation since those changes will manifest themselves primarily at
> those surfaces.
>
>> However in SIENAr there's a possibility to identify if a group
>> difference is
>> caused by atrophy of the first group or growth of the second one or
>> both
>> phenomena take place together. Is there any possibility in FSL-VBM to
>> disambiguate what's going on in areas identified with group
>> differences?The
>> statistics produced after vbm analysis only depict (the areas and)
>> the
>> significance of differences among the groups but not what actually
>> goes on
>> there.
>
> That is not correct. FSL-VBM is a more generic tool than SIENAr, and
> can be set up to answer much the same questions. For SIENAr you will
> need two time-points per subject, which is what allows you to compare
> changes-over-time between groups. If you have the same data (i.e. two
> time points per subject) you can set up an FSL-VBM analysis to look
> for time-by-group-interaction, which will pose much the same question
> as SIENAr does.
>
> Also, would this really be an issue? Are there any pathologies out
> there that causes sprouting of new brain tissue?
>
> Good Luck Jesper
|