Hi,
If you only have one subject then you can just use the results from the single run on that single subject from your first-level analysis. If you have multiple subjects and a single run per subject then you simply set-up a higher-level analysis that takes the first-level FEAT directories as input. You don't need to specifically extract any files yourself from the FEAT directory. See examples of standard higher-level analyses such as this in the FSL Course practicals.
All the best,
Mark
On 4 Jun 2014, at 15:40, L. Nestor <[log in to unmask]> wrote:
> I sent an email some time ago (see below) regarding one functional run of an fMRI paradigm. Steve Smith replied by saying that I could go straight to ME cross-subject analysis at the second level. Can I just confirm that the data that go into this analysis come from the reg_standard/stats directory (which I have for each subject)?
>
> Also, can somebody point me to a link/previous thread detailing how one can still perform a Fixed Effects analysis when you have only one run?
>
> Thanks,
>
> Liam.
>
> Hi - only having a single run per subject is actually probably the most common design and in this case you can go straight to ME cross-subject analysis at the second level.
> Steve
>
> On 11 Jun 2013, at 02:55, "L. Nestor" <[log in to unmask]> wrote:
>
>> Dear FSL users
>> What do you do in FSL when you only have one functional run of a task? Normally, you would have at least two functional runs, which you then average across in a higher level analysis using a Fixed Effects Model for each individual subject.
>> Therefore, how do you do a Fixed Effects analysis on a single subject in FSL when you have only one feat directory from your first level analyses on just one run?
>> Thanks,
>> Liam.
|