Dear FSL users,
I'm running a fixed-effects analysis in order to combining multiple runs from multiple subjects. I've performed a first-level analysis in each run separately. Registration and individual runs results seem to be ok. But, the results of the second-level analysis have some blobs outside the standard brain. The problem is that the mask created in the second level is bigger than the template. I've tried both linear and non linear registration in the first-level analysis.
Did someone have this issue before? Here the link for the images (summary of registration of each run (for both linear and non linear), masks (red: linear, yellow: non linear), and an example of results:
https://www.dropbox.com/sh/vtz5glvtc87pl86/AABzfcnIXRa-I0gAHPD_9q60a?dl=0
Besides, to me it seems that the linear registration works better than the non linear. Is this normal?
Thanks in advance!
Fernanda
|