 |
 |
Dear Martha, What do you mean when you say the files "differ"? They have different names? The contents are very different (e.g. 10% different)? Or the contents are different in the 9th decimal place? All the best, Mark On 7 Dec 2011, at 20:53, Martha Shiell wrote: > Hello, > > My question is regarding registration in a first-level FEAT analysis, when it is done for a single subject input versus for multi-subject inputs. > > When I run a FEAT analysis for a group of subjects who have the same matrix design, I set up the analysis once and include all of the subjects as input. Under the registration tab, I select "Main Structural Image" and input each subject's corresponding amri_brain image (I selected BET on the Pre-Stats tab, and the non-BET amri images are available in the same directory as the amri_brain images). I also select "Standard space", check off "non-linear", and input MNI152_T1_2mm_brain. > > When I run a FEAT analysis for a group of subjects who have different design matrices (for example, if I include a subject-specific regressor in the model), I set-up the registration identically to the description that I have given above, except that I only include my single subject's data as input, and only that subject's corresponding amri_brain on the Registration tab "main structural Image" input. > > I have noticed that some files in the reg folder of my .feat outputs differ between the two instances described above. The files that differ include all of the highres2standard* and example_func2standard*. However, the example_func2highres* files are identical between the two analyses. > > As I am using the same input functional data, the same highres T1, the same standard space image, and the same registration parameters in both cases, I don't understand why these two registrations are done differently. Please advise. > > Thank you, > Martha Shiell >