Re: [Marsbar-users] Timecourse and VOI for PPI questions

[Matthew B. <mat...@gm...>]

Hi,

Thanks for posting...

On Sun, Sep 18, 2011 at 7:58 AM, april unruh <apr...@sb...> wrote:
> Hi,
> I have a few questions:
> 1. I found the tutorial on how to extract raw timecourse data for specific
> ROIs from images but is there a way to extract this information by
> condition? For example, sort the data by fear face, calm faces...etc

Ah - no - sorry.  The problem is that you have to work out which time
points are from which condition.  That would be more or less OK for
simple block designs with long gaps between the blocks, but any other
design and it becomes  difficult to decide which time point goes with
which condition.   You could hack it by loading the design in matlab,
choosing all time points where a certain regressor was above (or below
- the undershoot) baseline, and get the time course from those time
points, but marsbar doesn't do that.

> 2. Is there a way to extract the time series information for all voxels in
> the brain?

Well - if you used the analysis brain mask as an ROI (by building an
ROI from the 'mask.img') - you could do this, but you might well run
out of memory.

> 3. I am going to also try running a PPI but I would like to create a VOI for
> each subject outside of SPM so that I can use a specific region of the brain
> and not an activated cluster. Is this possible? For example, I want the seed
> region to be the amygdala but since it is not activated in each subject I am
> unable to save a VOI within SPM.

I'm afraid marsbar doesn't currently read or write VOI's in SPM
format.  It's an obvious thing to do, I'm afraid that no-one (well, I
suppose particularly me) has done it yet.

> 4. Can the PPI be done by group?

Oh - dear - I don't know anything about PPI :)   Probably the SPM list
would be more help...

Best,

Matthew