Hi Cornelius,
if you send in the log file then we can have a look. If you're just
interested in identifying those components that show significant
correlation with the mc regressors you can also try to run fsl_glm on
melodic_mix against the design matrix
fsl_glm -i melodic_mix -d design.mat -c design.con -o output.txt
cheers
Christian
On 27 Jul 2008, at 21:41, Cornelius Werner wrote:
> Hi all,
>
> sorry if this is a trivial question.
>
> I am using the recent version of FSL on Ubuntu 8.04. I ran a resting
> state
> fMRI on 12 subjects, 216 volumes each. Running MELODIC in "temporal
> concatenation" mode, I noticed that there is quite some motion going
> on when
> browsing though the pre-stats graphics.
>
> What I wanted to do now is to include the mc parameters as EVs into
> melodic
> to identify those components that are most closely correlated with
> motion
> itself. That's what I did:
>
> 1) I ran mccutup on all relevant *.ica directories to obtain
> mc1.par...mc6.par files.
> 2) I concatenated mc1.par(subject1),
> mc1.par(subject2)...mc1.par(subject12)
> -> mc1_allsubjects.par
> 3) Repeated for mc2.par...mc6.par
> 4) In GLM-Setup I created a design matrix with the six large files
> from
> above as EVs 1-6 (custom EV, one entry)
> 5) Contrasts were 1 0 0 0 0 0, 0 1 0 0 0 0, ... , 0 0 0 0 0 1
> 6) F-Test included all six contrasts
> 7) Saved everything
>
> SO. I then included the resulting *.mat and *.con in MELODIC,
> started it
> again and waited. But: the output looks just the same?! I seem to be
> unable
> to find the additional information concerning the mc regressors etc.,
> neither in the output html nor in the folder (full stats).
>
> What was I doing wrong? Where should I look for the results from the
> GLM matrix?
>
> Thanks a lot in advance!
> Cornelius Werner
|