 |
 |
Hi Lorena, I do not know the details, but I believe that when you add the motion parameters, they are modified by filtering/detrending and/or by orthogonalization with respect to the rest of the design, and this will not necessarily happen when you add the raw motion parameters from the .par file in as confound regressors. You can look at the columns of the design.mat file and see what the motion parameters look like after whatever processing FSL does, and compare them to what you see in the .par file, they will be different. Perhaps someone at FSL can clarify the exact processing that occurs behind the scenes. Dan On Wed, Jul 25, 2012 at 3:03 PM, Lorena Jimenez-Castro <[log in to unmask]>wrote: > Hello, > > If I have a fmri data which has already been preprocessed successfully > in FEAT, and I want now to use FEAT for the remainder of the 1st level > analysis and include the motion parameters. Can I just run the Stats + > Post-stats using the filtered_func_data.nii image as input and add the > motion parameters by clicking the "add additional confound EVs" button in > the Feat GUI loading in the "prefiltered_func_data_mcf.par" file from the > original feat directory? Do I get the same results as if I run prestats, > stats and post_stats in the same FEAT run using "add motion parameters to > the model" button. > > I already tried and I got different results, So I was wondering why?? and > I would like to confirm that the results should be the same. If not, why > not? > > Thanks a lot > > Lorena >