Hi
On 26 Jun 2007, at 04:34, Xuelin Cui wrote:
> Dear all:
>
> I ran MELODIC 2 times for the same data. The 2 times were almost
> same parameter setting except for the FWHM value. The first time
> was 3 and the second time was 9. But the 2 results were different.
They should be different and is that not what you were aiming for?
> For the first time I got 42 ICs, while I had 73 ICs for the sencond
> time. I think the difference in number of components are not
> neglegible. Why wass that?
Smoothing changes the Eigenspectrum a lot - essentially it makes it
more lumpy, even in the minor part of the spectrum. The current
implementation of the dim.-estimation assumes signal in Gaussian
noise (non-smooth). It is possible to derive similar estimators for
a 'signal in smooth Gaussian noise' model but that's not part of the
current release version.
> is that simply because I changed the FWHM value from 3 to 9?
Well, normal run-to-run variability (keeping all parameters the same)
is low, so yes - it's related to all the things you changed between
runs, in this case your smoothing.
> How big impact will the FWHM value impose on the results?
>
Difficult to say beyond the trivial - I typically recommend smoothing
by about one voxel size (i.e. with 3mm voxels I typically stick with
3-5mm smoothing)
cheers
c
> Thanks
>
> Xuelin
>
> ****************************************
> Xuelin Cui
> Department of Electrical Engineering
> University of Hawaii-Manoa
> Honolulu HI 96822
>
> Tel: 1-808-349-0983
> Email: [log in to unmask]
> ****************************************
____
Christian F. Beckmann
University Research Lecturer
Oxford University Centre for Functional MRI of the Brain (FMRIB)
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK.
[log in to unmask] http://www.fmrib.ox.ac.uk/~beckmann
tel: +44 1865 222551 fax: +44 1865 222717
|