Hi - looking at the varcope input to second-level the second and last
timepoints have very high first-level variance. So you should look into
the first-level analyses for these subjects to test for bad outliers in
the first level timeseries. It may be that they were just very noisy, in
which case you probably want to just go with what you have, the final
second-level results can be trusted.
Cheers, Steve.
On Sat, 17 Dec 2005, Cherif Sahyoun wrote:
> Hi!
>
> Ok, the data, now with the design corrected, is uploaded under 647061.
>
> I went back and re-did everything in case there might still be a user error
> from my first analysis, and it seems fine to me (although you never know!).
> Just so you know, I had to tweak my low-level EVs because of an acquisition
> problem, but if that was to create a problem, i guess the low-level output
> would have complained...
>
> thx!
> c
>
> On 12/16/05, Stephen Smith <[log in to unmask]> wrote:
> >
> > Hi - sure, re-upload once you're done.
> > Cheers.
> >
> >
> > On Thu, 15 Dec 2005, Cherif Sahyoun wrote:
> >
> > > Hey Steve!
> > >
> > > My bad. Dunno how I let that through...
> > > I fixed it and started to rerun the analysis, and the same errors are
> > still
> > > showing up!
> > >
> > > Shall I upload the directory again when it's done?
> > >
> > > best,
> > > cherif
> > >
> > > On 12/14/05, Stephen Smith <[log in to unmask]> wrote:
> > > >
> > > > Hi Cherif - user-error I'm afraid ;-)
> > > >
> > > > Your first timepoint isn't being modelled - hence the model-fitting is
> > > > pretty unhappy! If you don't want to model the first timepoint then
> > just
> > > > exclude it from the analysis, or give it it's own EV.
> > > >
> > > > Cheers, Steve.
> > > >
> > > >
> > > >
> > > >
> > > > On Wed, 14 Dec 2005, Cherif Sahyoun wrote:
> > > >
> > > > > Hey Steve!
> > > > >
> > > > > Thx!
> > > > > It's uploaded with ID 904366 as grp_cope_better.gfeat.tar.gz
> > > > >
> > > > > best,
> > > > > cherif
> > > > >
> > > > > On 12/13/05, Stephen Smith <[log in to unmask]> wrote:
> > > > > >
> > > > > > Hi - in this case the outliers may be subtle - if you want to send
> > > > your
> > > > > > data (the complete .gfeat directory) we can have a quick look:
> > > > > >
> > > > > > Please upload the files in a single compressed tarfile to
> > > > > > http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi
> > > > > > and then email me the upload ID.
> > > > > >
> > > > > > Cheers.
> > > > > >
> > > > > >
> > > > > > On Mon, 12 Dec 2005, Cherif Sahyoun wrote:
> > > > > >
> > > > > > > Hey Steve!
> > > > > > >
> > > > > > > done that. no luck... no ouliers outside the brain, and looking
> > at
> > > > the
> > > > > > > timecourses, no voxel seems to be much different from its
> > > > neighbours...
> > > > > > > Anything in particular I am looking for? I may be looking at the
> > > > file
> > > > > > wrong,
> > > > > > > but it looks fine to me... what else could the problem be?
> > > > > > >
> > > > > > > thx,
> > > > > > > cherif
> > > > > > >
> > > > > > > On 12/12/05, Stephen Smith <[log in to unmask]> wrote:
> > > > > > > >
> > > > > > > > Hi - you should probably set the colour lookuptable to
> > Greyscale
> > > > and
> > > > > > the
> > > > > > > > display range to whatever
> > > > > > > > avwstats++ <fmridata> -r
> > > > > > > > shows. Then flick through the timepoints and also turn on
> > > > timeseries
> > > > > > > > viewing in order to see strong outliers easily.
> > > > > > > >
> > > > > > > > Cheers.
> > > > > > > >
> > > > > > > >
> > > > > > > > On Sun, 11 Dec 2005, Cherif Sahyoun wrote:
> > > > > > > >
> > > > > > > > > Hey Steve!
> > > > > > > > >
> > > > > > > > > Ok, I opened a bunch of copes in fslview. I don't have other
> > > > data
> > > > > > handy,
> > > > > > > > so
> > > > > > > > > i don't know if that's normal, but the background for all
> > copes
> > > > is a
> > > > > > > > uniform
> > > > > > > > > "orange", along the same tones as inside the brain... should
> > it
> > > > be
> > > > > > > > "black"
> > > > > > > > > i.e. nothing there compared with the brain (which goes from
> > > > > > > > > yellow-orange-red)?
> > > > > > > > > Also, on a few of the copes, on the edge of the brain
> > usually,
> > > > there
> > > > > > are
> > > > > > > > a
> > > > > > > > > couple of point that go down to "black" (individual pixels,
> > > > usually
> > > > > > only
> > > > > > > > > 1/cope image).
> > > > > > > > >
> > > > > > > > > Does that help hone in on the problem?
> > > > > > > > >
> > > > > > > > > thx,
> > > > > > > > > cherif
> > > > > > > > >
> > > > > > > > > On 12/10/05, Stephen Smith <[log in to unmask]> wrote:
> > > > > > > > > >
> > > > > > > > > > Hi - no, ignore the FLIRT warnings.
> > > > > > > > > >
> > > > > > > > > > The FLAME warnings are of more concern - this almost
> > always
> > > > means
> > > > > > that
> > > > > > > > you
> > > > > > > > > > have very strong outliers in your data input to FLAME. One
> > > > thing
> > > > > > to
> > > > > > > > check
> > > > > > > > > > is to load the .gfeat/cope? images into fslview and flick
> > > > through
> > > > > > the
> > > > > > > > > > timepoints to check for outliers.
> > > > > > > > > >
> > > > > > > > > > Cheers, Steve.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > On Fri, 9 Dec 2005, Cherif Sahyoun wrote:
> > > > > > > > > >
> > > > > > > > > > > Hi!
> > > > > > > > > > >
> > > > > > > > > > > 2 errors I am getting when running high level analyses
> > in
> > > > Feat:
> > > > > > > > > > > - same as in my previous email, everytime flirt is
> > called, I
> > > > get
> > > > > > an
> > > > > > > > > > error
> > > > > > > > > > > stating
> > > > > > > > > > > WARNING: Both reference and input images have an sform
> > > > matrix
> > > > > > set
> > > > > > > > > > > The output image will use the transformed sform from the
> > > > > > reference
> > > > > > > > image
> > > > > > > > > > > This happens both in low level and high level analyses
> > (more
> > > > > > details
> > > > > > > > in
> > > > > > > > > > my
> > > > > > > > > > > previous message)
> > > > > > > > > > >
> > > > > > > > > > > the reason I am repeating this is it may be related to
> > the
> > > > > > following
> > > > > > > > > > > warnings I get in high level stats:
> > > > > > > > > > >
> > > > > > > > > > > WARNING: FLAME stage 2 has given an abnormally large
> > > > difference
> > > > > > to
> > > > > > > > stage
> > > > > > > > > > 1
> > > > > > > > > > > x=50,y=92,z=41
> > > > > > > > > > > vox=173818
> > > > > > > > > > > zts[0](vox)=7.7325
> > > > > > > > > > > zemlowerts[0](vox)=4.62063
> > > > > > > > > > > beta_mean[0](vox)=2.73318e-05
> > > > > > > > > > > beta_b[0](vox)/beta_c[0](vox)=1.00022e-10
> > > > > > > > > > >
> > > > > > > > > > > WARNING: FLAME stage 2 has given an abnormally large
> > > > difference
> > > > > > to
> > > > > > > > stage
> > > > > > > > > > 1
> > > > > > > > > > > x=50,y=92,z=42
> > > > > > > > > > > vox=173819
> > > > > > > > > > > zts[0](vox)=7.99212
> > > > > > > > > > > zemlowerts[0](vox)=4.7153
> > > > > > > > > > > beta_mean[0](vox)=2.73318e-05
> > > > > > > > > > > beta_b[0](vox)/beta_c[0](vox)=1.00022e-10
> > > > > > > > > > >
> > > > > > > > > > > I get MANY iterations like these before the analysis
> > finally
> > > > > > ends
> > > > > > > > (in
> > > > > > > > > > > >24hrs!), and am not sure at all what this means or if
> > it's
> > > > > > > > normal...
> > > > > > > > > > >
> > > > > > > > > > > any views?
> > > > > > > > > > >
> > > > > > > > > > > thx!
> > > > > > > > > > > cherif
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > --
> > > > > > > > > > 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
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > --
> > > > > > > > >
> > > > > > > >
> > > > > >
> > > >
> > -----------------------------------------------------------------------------------------
> > > > > > > > > Cherif P. Sahyoun
> > > > > > HST-MEMP
> > > > > > > > >
> > > > > > > > > Developmental Neuroimaging of Cognitive Functions
> > > > > > > > >
> > > > > > > > > C: 617 688 8048
> > > > > > > > > H: 617 424 6956
> > > > > > > > > [log in to unmask]
> > > > > > > > >
> > > > > > > > > "Live as if this were your last day. Learn as if you'll live
> > > > > > forever"
> > > > > > > > > Ghandi
> > > > > > > > >
> > > > > > > >
> > > > > >
> > > >
> > -------------------------------------------------------------------------------------------
> > > > > > > > >
> > > > > > > >
> > > > > > > > --
> > > > > > > > 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
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > --
> > > > > > >
> > > > > >
> > > >
> > -----------------------------------------------------------------------------------------
> > > > > > > Cherif P. Sahyoun
> > > > HST-MEMP
> > > > > > >
> > > > > > > Developmental Neuroimaging of Cognitive Functions
> > > > > > >
> > > > > > > C: 617 688 8048
> > > > > > > H: 617 424 6956
> > > > > > > [log in to unmask]
> > > > > > >
> > > > > > > "Live as if this were your last day. Learn as if you'll live
> > > > forever"
> > > > > > > Ghandi
> > > > > > >
> > > > > >
> > > >
> > -------------------------------------------------------------------------------------------
> > > > > > >
> > > > > >
> > > > > > --
> > > > > > 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
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > > --
> > > > >
> > > >
> > -----------------------------------------------------------------------------------------
> > > > > Cherif P. Sahyoun
> > HST-MEMP
> > > > >
> > > > > Developmental Neuroimaging of Cognitive Functions
> > > > >
> > > > > C: 617 688 8048
> > > > > H: 617 424 6956
> > > > > [log in to unmask]
> > > > >
> > > > > "Live as if this were your last day. Learn as if you'll live
> > forever"
> > > > > Ghandi
> > > > >
> > > >
> > -------------------------------------------------------------------------------------------
> > > > >
> > > >
> > > > --
> > > > 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
> > > >
> > >
> > >
> > >
> > > --
> > >
> > -----------------------------------------------------------------------------------------
> > > Cherif P. Sahyoun HST-MEMP
> > >
> > > Developmental Neuroimaging of Cognitive Functions
> > >
> > > C: 617 688 8048
> > > H: 617 424 6956
> > > [log in to unmask]
> > >
> > > "Live as if this were your last day. Learn as if you'll live forever"
> > > Ghandi
> > >
> > -------------------------------------------------------------------------------------------
> > >
> >
> > --
> > 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
> >
>
>
>
> --
> -----------------------------------------------------------------------------------------
> Cherif P. Sahyoun HST-MEMP
>
> Developmental Neuroimaging of Cognitive Functions
>
> C: 617 688 8048
> H: 617 424 6956
> [log in to unmask]
>
> "Live as if this were your last day. Learn as if you'll live forever"
> Ghandi
> -------------------------------------------------------------------------------------------
>
--
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, 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
|