Hi Steve,
I'd like to know how to compute an F-Test in an higher level analysis
(specific examples, shouldn't make a difference though)): 1) Single
Subject, multiple runs (say 8), two EVs, would like to get F-test ( 1 0
, 0 1) across runs 2) Single Subject 2 real EVs modeled with a set of
Gamma basis functions, multiple runs, 3) How to bring F-test results
from 1 and 2 to the 3rd level (across subjects).
Thanks,
h.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
Behalf Of Stephen Smith
Sent: Thursday, November 21, 2002 10:46 AM
To: [log in to unmask]
Subject: Re: [FSL] Single Subject multi session - F-Test?
Hi Hauke - exactly what two-session design did you want to do an F-test
for?
thanks, Steve.
On Tue, 19 Nov 2002, Hauke R. Heekeren wrote:
> Hi there,
> Along those lines: how can we compute an F-Test for these two runs?
>
> Thanks,
>
> h.
>
> -----Original Message-----
> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
> Behalf Of Stephen Smith
> Sent: Wednesday, November 13, 2002 9:49 AM
> To: [log in to unmask]
> Subject: Re: [FSL] Single Subject multi session
>
> Hi - you are right in saying that if you want to combine just two runs
> as
> if they had been one then this should be approached as a fixed-effects
> analysis. In this case you want to use simple command-line combination
> of
> the cope and varcope images in the <featdir>/stats subdirectories; for
> example:
>
> avwmerge -t COPES <featdir1>/stats/cope1 <featdir2>/stats/cope1
> avwmerge -t VARCOPES <featdir1>/stats/varcope1
<featdir2>/stats/varcope1
>
> avwmaths VARCOPES -Tmin MASK
> avwmaths VARCOPES -Tmean -div 2 -mas MASK VARCOPE
> avwmaths COPES -Tmean COPE
>
> ttoz -zout FIXEDZ VARCOPE COPE <DOF>
>
> (where <DOF> is the sum of the values in the different stats/dof files
> in
> the different FEAT directories)
>
> You can then do normal thresholding for example using easythresh.
>
> Note - the above assumes that the two runs are registered to each
other
> -
> which will not in general be exactly true for first-level analyses.
The
> easy thing to do if you have run FEAT registration is to first run
> featregapply <featdir>
> on each FEAT directory and then use the resulting
> copes and varcopes from the <featdir>/reg_standard/stats
subdirectories
> instead of just the stats subdirectory.
>
> Thanks, Steve.
>
>
> On Wed, 13 Nov 2002, Thomas Stephan wrote:
>
> > Dear FSL-List,
> >
> > how can I perform a fixed effects analysis of two fmri imaging runs
> acquired
> > from one subject
> > in one day ? It is easy to do the analysis of one run using FEAT,
but
> I
> > could not find out
> > how to enter two runs at once.
> >
> > Thank you for your support
> > Thomas Stephan
> >
======================================================================
> > Thomas Stephan Email:
[log in to unmask]
> > Klinikum Grosshadern
> > Neurologisches Forschungshaus
> > Marchioninistr. 23 Fon: +49 089 / 7095-4819
> > 81377 Muenchen Fax: +49 089 / 7095-4801
> >
======================================================================
> >
>
> Stephen M. Smith
> Head of Image Analysis, FMRIB
>
> Oxford University Centre for Functional MRI of the Brain
> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> +44 (0) 1865 222726 (fax 222717)
>
> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
>
Stephen M. Smith
Head of Image Analysis, FMRIB
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|