Hi Anderson,
I highly appreciate your guidance!
My seed is a statistical map from previous PET analysis between the groups in the same cohort.
The statistical map was generated using "Randomise" then I thresholded it at 0.99 to create the seed of interest. This seed is only one volume and it is a 3D image.
I am not fully understand what you mean by "In order to make a single 4D file (one volume per seed/mask)". Kindly, do I need to create multiple copies of the 3D seed ( copy per subject) then merge using "fslmerge and the flag -t" to get the 4D mask?
Thank you
Jon
Hi Jon,
I like using the dual regression. You'd supply the masks with the seeds (non-overlapping) as a single 4D file (one volume per seed/mask), in the place of what usually would be the melodic_IC.
All the best,
Anderson
On 6 October 2016 at 18:32, John anderson <[log in to unmask]> wrote:
Dear Experts,
Please forgive the simplicity of my question. I have resting state data. I did the preprocessing steps in FEAT and I got the filtered functional data.
I want to do seed based correlations analysis.
I am wondering to which script I must feed to the filtered func data. To Fsl_glm or to dual_regression. My seed is a 3D ROI represent the difference between the subjects in PET data.
Thank you for any suggestion
Jon
|