 |
 |
Hi, You could avoid moving data into the standard space in step 3 and instead move the masks from the standard space into the native space. Then you can things in the native space. However, if you are going to analyse the timeseries across different sessions at the group level, then you will need to transform into the standard space anyway at some point so it probably doesn't make much difference. As for temporal derivatives, I would say that it might be helpful if the TR was larger, in order to accommodate changes in the slice timing. For filtering, it depends if your extracted timeseries have already been filtered prior to being extracted. If they have then you need to avoid filtering twice. If not then you must filter so that it matches what is being applied to the data. All the best, Mark On 23 May 2014, at 16:04, Kris Farrant <[log in to unmask]> wrote: > Hi Mark sorry for the late reply. > > I am doing resting state data analysis. I am using ABIDE data that is publicly available (the download is 2 files. 1 anatomical (Mprage) and 1 functional (rest)). I thought I overcome this problem with the following steps: > > 1) BET extract Mprage > 2) Run Feat > 3) Convert filtered_func file into MNI 152 standard space using Flirt option > 4) Generate ROI which is made in standard 152 space > 5) Extract time course from this ROI using my newly generated standard_space_filtered_func file > 6) Convert the Mprage_brain into standard space > 7) FAST segmentation using standard space Mprage brain and standard space tilt fund file. > 8) Extract time course of the White matter and CSF into .txt files > 9) use the CSF, WM time course .txt files and input them as EVS to regress out of my ROI time course. I have selected orthogonalise, temporal derivative and temporal filter for the ROI EV but not for the two nuisance regressors. > 10) Use the newly generated .feat file in a group level analysis. > > My main concerns are: > 1) Will transforming my data into standard space like this change my data? > 2) Is it correct to select the temporal derivative and temporal filter options in step 9? > > Could you please tell me if this pipeline seems reasonable or not? > > Thanks