Hi,
Run it both ways, and you should see that the results are the same either
way. And if you accounted for the degrees of freedom removed with
approach (2), the statistics should actually be identical.
cheers,
-MH
--
Michael Harms, Ph.D.
-----------------------------------------------------------
Conte Center for the Neuroscience of Mental Disorders
Washington University School of Medicine
Department of Psychiatry, Box 8134
660 South Euclid Ave.Tel: 314-747-6173
St. Louis, MO 63110Email: [log in to unmask]
On 7/29/15 9:04 AM, "Andrew Song" <[log in to unmask]> wrote:
Dear FSL experts,
It is my understanding that fsl_regfilt can also be used to regress out
nuisance factors such as CSF and WM.
In this regard, when running seed connectivity analysis, I was wondering
which method is preferable, or whether they have any difference at all.
1) Run fsl_glm on original data with extracted time-series (seed) and
nuisance regressors (motion parameters, CSF, WM).
2) Using fsl_regfilt, remove nuisance regressors from original data, and
then simply run glm on cleaned data with only the seed time-series in the
design matrix.
My general approach when extracting time-series from the seed region is to
regress out nuisance factors from the seed time-series first (using
fsl_regfilt) and then use fslmaths to extract the time-series that is
ideally unaffected by non-neural factors. With this raw signal, I was not
sure which of the two methods to pursue.
Thanks,
________________________________
The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.
|