Hi Darren, the relevant part of the manual is at:
http://www.fmrib.ox.ac.uk/fsl/feat5/detail.html#tsplot
but yes I'm afraid there isn't much text there on how to interpret the
plots. In general you expect the blue lines to reflect your model, fitted
to the data, so these should actually look like the model that you put in.
We'll try to expand on the text a little for the next release.
Cheers.
On Thu, 4 Mar 2004, Darren Schreiber wrote:
> This post was helpful to me. Is there more information somewhere on
> the FSL website about how to read the time series plots or the
> peri-stimulus plots? In particular, how to detect a failure of
> analysis or other things to look for with them on the individual and
> higher levels?
>
> It might even be helpful to build "FSL help page" links into the
> report.html files?
>
> Darren
>
>
>
> On Mar 2, 2004, at 5:54 AM, Mark Jenkinson wrote:
>
> > Hi Jane,
> >
> > This sounds like you have done everything correctly.
> > The "time series" plots that are generated at the higher level
> > actually reflect the parameter estimates (copes) across
> > session/subject, not across time. So you expect to see
> > some form of zig-zag shape as this reflects the variability
> > of the response from session to session (or subject to
> > subject if that's what your higher level analysis is across).
> >
> > If you are getting sensible activations and there are no
> > obvious "outliers" in these "time series" plots (which could
> > indicate a potential failure of one of the individual session
> > analyses) then it sounds like everything is fine.
> >
> > All the best,
> > Mark
> >
> >
> > Jane Aspell wrote:
> >
> >> hi
> >> thanks for the advice with my previous problems. this one kind of
> >> follows
> >> on...the time series plots for my higher level analysis (i was
> >> averaging
> >> together the results of 5 identical runs/series) look really bad. the
> >> full
> >> model and cope model fits are both straight lines and my data is a
> >> kind of
> >> zig zag shape!
> >> is this because of how i set up my EVs and contrasts?...i had one EV
> >> and one
> >> group (as i wanted to get the average over the runs for each of my 4
> >> contrasts ) and so i had one column in the design matrix with ones in
> >> each
> >> element.
> >> so in other words - is it ok that the time series look like this or
> >> has
> >> something gone wrong? the activations look fine and the analysis seems
> >> otherwise to have worked.
> >>
> >> thanks again
> >>
> >> jane
> >>
> >>
> >
>
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
|