Hi - I'm afraid there is still no general agreement on the 'optimal'
approach to such confounds. This is not specific to any particular
software (FSL, SPM etc) but a generic question relating to what best
to put in your GLM. If you have a good feel for data and results you
might try both ways and see which makes more sense to you.
If you run MELODIC using the filtered_func_data in the FEAT output
directory then the data will by default have already been betted. If
you clean up that data using MELODIC you should then feed that into a
further FEAT analysis with preprocessing turned off.
Cheers, Steve.
On 27 Mar 2008, at 19:32, Carlos Faraco wrote:
> Dear FSL users,
>
> I just wanted some input on whether adding motion parameters to the
> model is
> now something that is widely accepted. In the FSL manual it cautions
> against
> this, stating that not much is known about what this does to the data.
>
> If adding these motion parameters is now something that is widely
> accepted
> for FSL, what produces more accurate removal of motion artifacts,
> removing
> motion through MELODIC or adding the motion parameters.
>
> Also, I have tried looking for guidelines on how to identify motion in
> MELODIC, and have only found some posts stating that it will be seen
> at the
> edge of the brain. I personally feel uncomfortable with such a vague
> description. Is there any other information available on how to do
> this? I
> have looked through the FMRIB Centre website for information, but I
> haven't
> found anything there.
>
> Lastly, when running MELODIC to identify motion, should the images
> be BETed?
>
> Thank you,
>
> Carlos
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|