Dear Adam,
> Hi all,
>
> Two Questions:
>
> 1. In my lower level analyses I would like to get rid of some time points that
> have excessive motion. Online I saw I could create a dummy EV. Also I was
> thinking I could just omit the time by changing the specifics of my 3-column
> format. Will both work? Is there any other way I could do it?
You will not be able to omit the time using the 3-column format. The
3-col format just allows for specification of stimulus/regression timings.
The best way to do this is as the web pages suggests which is to use a
dummy EV.
> 2. I tried to use FLAME for a higher level analysis among runs within the same
> subject. For some reason the analyses would not be completed (I received
> results for between 5-9 contrasts out of 31). However, FLAME for stage 1 only
> worked fine. I am planning to use FLAME for the next analysis across subjects
> but am worried it may not work. Any thoughts?
If you are going to use pass up the results of a within subject between
session analysis to a third level then you should only do FLAME stage 1
anyway. This is because FLAME stage 2 would need to be run on every voxel
(rather than on those near to threshold - as is normally done) to get
reliable results, and this is computationally too expensive. As for
getting no results for certain contrasts, it depends what you mean, it may
be that there are no voxels/clusters passing the threshold.
Cheers, Mark.
Mark Woolrich.
Oxford University Centre for Functional MRI of the Brain (FMRIB),
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK.
Tel: (+44)1865-222782 Homepage: http://www.fmrib.ox.ac.uk/~woolrich
|