OK - I think I know what the difference is - it's probably in the
before-t=0 padding of the data that happens in order to allow the
convolution to work. I can't remember what the padding rule is right now -
but it may just be zeros - which would explain this slight difference at
the very beginning after convolution.
Hope that all makes sense!
ttfn :)
On Wed, 29 May 2002, lino becerra wrote:
> Hi Steve:
> I figured out the difference (yet I quite don't understand it).
> My custom file was filled with 1's and -1's (baseline=-1). The main difference was at
> the beginning and end of the model. The custom file-based model showed
> steeper changes in the initial and final baseline. However, If I shifted the custom file
> to 0's and 1's (i.e., baseline=0) I can reproduce the model created using the Square
> option. By playing with the filter and convolution options I found that it was the convolution
> causing the differences. I don't understand why the convolution of the gamma function with a
> shifted function produces different results.
> thanks for your help
> lino
>
> Stephen Smith wrote:
>
> > HI - yes, they will be the same - in what ways are they differing for you?
> > if you send the exact AB parameters and the custom file we should be able
> > to tell you what's causing the difference.
> >
> > Thanks, Steve.
> >
> > On Tue, 28 May 2002, Lino Becerra wrote:
> >
> > > Hi:
> > > I tried the following experiment :
> > > A.- I used the Basic shape Square option when defining EV1 under the Full model setup with 30 s off/25 on, TR=2.5 , 100 volumes.
> > > B.- Then I used the Custom (1 entry per volume) Basic shape option. The file I used in this case
> > > was the "same" as the oneyou might get from the Square option; I generated a text file of 1's and -1's
> > > to emulate the on/off conditions. I kept everything else the same (i.e. the Advanced options).
> > > My thought was that the resulting models under A and B should be the same, but they weren't.
> > >
> > > Do you know why?
> > > Thanks
> > > lino
> > >
> >
> > Stephen M. Smith
> > Head of Image Analysis, FMRIB
> >
> > 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
>
> --
> Lino Becerra PhD
> Assistant Professor
> Martinos Center For Biomedical Imaging
> 149 13th St
> Charlestown, MA 02129
> P: 617-726-4231
> F: 617-724-7495
>
Stephen M. Smith
Head of Image Analysis, FMRIB
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
|