Hi - that sounds right - sounds like the various options for higher-
level models in the FEAT manual should cover what you need - so I
recommend you have a look there.
Cheers, Steve.
On 9 Jul 2009, at 10:57, Hallvard Røe Evensmoen wrote:
> Thanks for the answer, you were right on the spot.
> I now wonder if anyone have any suggestions about how I should set
> up my analysis. We have two groups of 20 subjects (patients and
> controls) who have performed four tasks during scanning (task1,
> task2, task3, task4 scanned as four different runs). We wanted to do
> the 2nd level analysis by comparing "task1 vs task2"-contrast,
> "task1 vs task3"-contrast etc. in each subject. In the 3rd level
> analysis we wanted to compare the patient group's average contrast
> with the control group's average contrast. Are on the wrong track
> here, or should we just trick FSL by adding a fifth "dummy" run that
> isn't used later in the analysis?
>
>
> Best regards
>
> Hallvard Røe Evensmoen
>
> -----------------------------------------------------------------------------
> Siterer "David V. Smith" <[log in to unmask]>:
>
>> It's probably giving the error because you don't have any degrees of
>> freedom. It looks like your model has 4 observations and 4
>> parameters.
>> Presumably, your three-EV model should produce the same error if
>> you only
>> provide it with 3 runs worth of data.
>>
>> David
>>
>>
>>
>>
>> -----Original Message-----
>> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
>> Behalf
>> Of Hallvard R øe Evensmoen
>> Sent: Wednesday, July 08, 2009 9:44 AM
>> To: [log in to unmask]
>> Subject: [FSL] FEAT 2nd level fails when including an additional EV
>>
>> Data are from four runs in the same subject. First-level FEAT
>> analysis has
>> been done on the different runs. The error message shows up when
>> increasing
>> to four EVs in a second-level analysis.
>> It works smoothly on three EVS, and fails no matter what first-
>> level EPI we
>> use in the fourth EV (thus independent of subject or run). We got
>> the same
>> error message when running on a different computer with different
>> OS (OS X
>> vs Linux).
>> ---------------------------------------------------------------------
>> Erroe message:
>>
>> cat ../design.lcon | awk '{ print }' > design.lcon
>> /Applications/fsl/bin/flameo --cope=filtered_func_data
>> --vc=var_filtered_func_data --mask=mask --ld=stats --dm=design.mat
>> --cs=design.grp --tc=design.con --fc=design.fts --runmode=flame1
>>
>> An exception has been thrown
>> nevs >= ntpts, Singular matrix.Trace: Gsmanager::flame_stage1;
>> Gsmanager::run.
>> ---------------------------------------------------------------------
>> The model setup has been attatched.
>>
>> Does anyone know what might be wrong?
>>
>>
>>
>> Best regards
>>
>> Hallvard Røe Evensmoen
>>
>>
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|