Hi - this makes sense - the different sets of commands are because one of
the data sets is just big enough to cause FEAT to switch into calling
FLAME one slice at a time, so that the memory requirements do not get too
big. This doesn't make any difference to the results, as FLAME does not
contain spatial processes.
Note, however, that it is unlikely that you want to be doing this paired
test in two separate analyses - I would have expected you to want to be
processing both groups in one big model, as in the example on the web
page.
Cheers, Steve.
On Wed, 18 Aug 2004, Michal Kuniecki wrote:
> Hello everyone,
>
> I have problems running second level analisys. My setup is quite simple.
> Namely I have two groups (males and females) and each o them went through
> two sessions. I want to know if there is any activation present within the
> gender group due to a session. Therefore I guess that the two paired t test
> (one for males one for females) is appropriate. I prepared everthing just
> as it is in the online tutorial for paired t-test. However when I run the
> script for females in certain part of the analisys I get :
>
> ln -s ../tdof_t1.hdr tdof_filtered_func_data.hdr
>
> ln -s ../tdof_t1.img tdof_filtered_func_data.img
>
> /usr/local/fsl/bin/avwroi mask tmpmask 0 91 0 109 0 1
>
> /usr/local/fsl/bin/avwroi filtered_func_data tmpcope 0 91 0 109 0 1
>
> /usr/local/fsl/bin/avwroi var_filtered_func_data tmpvarcope 0 91 0 109 0 1
>
> /usr/local/fsl/bin/flame --cope=tmpcope --vc=tmpvarcope --mask=tmpmask --
> ld=stats0 --dm=design.mat --cs=design.grp --tc=design.con --ols --
> nj=10000 --bi=500 --se=1 --fm --zlt=2 --zut=20
> Log directory is: stats0
> Setting up:
> No f contrasts
> nevs=22
> ntpts=42
> ngs=1
> nvoxels=0
> Running:
> nmaskvoxels=0
> njumps = 10000
> burnin = 500
> sampleevery = 1
> nsamples = 9500 and than it creates following stats directories
>
> and for males I gest something completly else:
>
>
> ln -s ../tdof_t1.hdr tdof_filtered_func_data.hdr
>
> ln -s ../tdof_t1.img tdof_filtered_func_data.img
>
> /usr/local/fsl/bin/flame --cope=filtered_func_data --
> vc=var_filtered_func_data --mask=mask --ld=stats --dm=design.mat --
> cs=design.grp --tc=design.con --ols --nj=10000 --bi=500 --se=1 --fm --
> zlt=2 --zut=20
> Log directory is: stats
> Setting up:
> No f contrasts
> nevs=20
> ntpts=38
> ngs=1
> nvoxels=122082
> Running:
> nmaskvoxels=8187
> njumps = 10000
> burnin = 500
> sampleevery = 1
> nsamples = 9500
>
> so basically three lines are missing here:
>
> /usr/local/fsl/bin/avwroi mask tmpmask 0 91 0 109 0 1
>
> /usr/local/fsl/bin/avwroi filtered_func_data tmpcope 0 91 0 109 0 1
>
> /usr/local/fsl/bin/avwroi var_filtered_func_data tmpvarcope 0 91 0 109 0 1
>
>
> and also the flame parameters are completly different. Finally both
> analises create reports with activations but I suspect that something wrong
> is going on in case of analisys for males. Can you please verify I my
> suspicion is right ?
>
> With best regards
> Michal Kuniecki
>
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
|