Hi - I think you are saying that for each subject you have 8 first-level
sessions, 4 with rest and active and 4 with rest and neutral conditions.
If this is correct then you should analyse each first-level session
separately and combine everything in one big second-level analysis, at
which point you can make contrasts between active and neutral conditions.
Thanks, Steve.
On Wed, 14 Jan 2004, Philipp Prahs wrote:
> Hi,
>
> sorry if anything remains unclear here - my knowledge of neither fsl nor
> english is very advanced.
> At first level I used only one EV and one contrast ("1") according to
> our paradigm (40 seconds rest,
> 16 second active or neutral stimulation, repeated 3 times). I ran the first
> level analysis seperately for "active" and "neutral" conditions. Are you
> suggesting that I could
> run a first level analysis including both "active" and "neutral"
> conditions with 2 EVs and contrasts
> "1 0", "0 1", "1 -1" and "-1 1". I did not think of that in the first
> place because I have seperate files
> containing "active" and "neutral" conditions. But I think it should be
> possible to merge them with avwmerge.
>
> Thanks very much for any comments or suggestions.
>
> Philipp
>
>
> Stephen Smith wrote:
>
> >Hi - I'm struggling to understand exactly your question, because I'm not
> >sure what your first-level analysis was - what EVs and contrasts did you
> >run at first level?
> >
> >Thanks, Steve.
> >
> >
> >On Wed, 7 Jan 2004, Philipp Prahs wrote:
> >
> >
> >
> >>Hi,
> >>
> >>I am trying to analyze a couple of measurements of 4 subjects who are
> >>measured under an active and a neutral condition repeatedly (4 times). After
> >>carrying out the first level analysis seperately I started a higher level
> >>analysis containing all 4 subjects. I set it up with 4 EVs (one for each
> >>subject) and ran the analysis for the "active" and "neutral" condition
> >>seperatly.
> >> EV1 EV2 EV3 EV4
> >>Subj1 1st Meas 1 0 0 0
> >>Subj1 2nd Meas 1 0 0 0
> >>Subj1 3rd Meas 1 0 0 0
> >>Subj1 4th Meas 1 0 0 0
> >>Subj2 1st Meas 0 1 0 0
> >>.
> >>.
> >>.
> >>
> >>Contrasts
> >>Subj1 1 0 0 0
> >>Subj2 0 1 0 0
> >>Subj3 0 0 1 0
> >>Subj4 0 0 0 1
> >>
> >>After that I ran an analysis using the cope images from the analysis above
> >>to get the group mean with one ev and all ones in the matrix
> >>
> >>My first question: Is it now legal to use the resulting cope images of the
> >>"active" and "neutral" conditions carried out similarly to get a contrast of
> >>"active" more than "neutral", like to EVs one for each cope image and a
> >>contrast with 1 -1 and -1 1 for both directions?
> >>
> >>My second question: When I try to get that contrast in a higher level
> >>analysis of only one measurement of one subject including only the "active"
> >>and the "neutral" measurement (2 datapoints) with 2 EVs 1 0 (active) and 0 1
> >>(neutral) with contrasts 1 0 (active only) 0 1 (neutral only) 1 -1 (active
> >>more than neutral) and -1 1 (neutral more than active) flame complains about
> >>the matrix being singular. Where did I make a mistake? I thought that this
> >>analysis should be possible.
> >>
> >>Thanks very much for any help with any of those questions.
> >>
> >>Philipp Prahs
> >>Dept. of Radiology - LMU Munich
> >>
> >>
> >>
> >
> > Stephen M. Smith DPhil
> > Associate Director, FMRIB and Analysis Research Coordinator
> >
> > 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 DPhil
Associate Director, FMRIB and Analysis Research Coordinator
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
|