Hi Sam, thanks for sending the design. I presume that this is the second
level of a 3-level analysis. As it says in the bubble-help for FLAME, if
you are going to pass FLAME results up to a higher level (as you will do
here) then you should be running FLAME stage 1 only at the second level,
which should speed things up a lot for you.
I've amended the doc so that in the next release this 3-level example
reminds people about this - thanks!
Let me know if this doesn't fix things - cheers, Steve.
On Wed, 5 May 2004, Stephen Smith wrote:
> Hi - yes, this is unusually slow. You shouldn't have any problems with
> RAM, because you have >40 inputs at second level, so FLAME will be being
> run slice-by-slice. Could you send in design.gif from second level so we
> can have a quick look at that.
>
> In the meantime, you could try running FLAME (part 1 only) to check that
> reasonable results are coming out.
>
> Cheers, Steve.
>
>
> On Tue, 4 May 2004, Sam Harris wrote:
>
> > I have a question about how long a higher lever analysis can be
> > reasonably expected to run. I recently set up a second level analysis
> > on a dual G5 with 2.5 gigabytes of RAM, and after 36 hours it was
> > still struggling to finish the the first COPE. At this rate, the
> > analysis would have taken over two weeks to terminate. Is this even
> > in the range of possibility for a valid design? Or is the problem
> > necessarily in the brain of the user?
> >
> > The design came straight from the FSL documentation (Multi-Session &
> > Multi-Subject [Repeated Measures -- Three Level Analysis]). We ran an
> > event-related protocol on 14 subjects, 3 scans each, 170 volumes per
> > scan. There were 12 contrasts done at the lower level. FLAME was run
> > at both levels.
> >
> > Thanks in advance for your help.
> > Best,
> > Sam
> >
>
> 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
>
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
|