thanks a lot for your reply. we did figure out that coreg failed due to FOV took off few slices on the top. Also, 'register to mean' rather than 'register to first' makes registration better. Plus severe motion added huge displacement of voxels and artifacts
Kushal J Kapse
Aphasia Research Laboratory
Boston University
________________________________________
From: H. Nebl [[log in to unmask]]
Sent: Thursday, January 09, 2014 11:44 AM
To: [log in to unmask]; Kapse, Kushal Janardan
Subject: Re: Activation on skull
I haven't seen the pictures, but activations on the skull point to motion artefacts. This would be compatible with "Its in mean realigned image and thereafter which seems to cutoff the top brain". As the mean is based on those parts of the brain covered / "not empty" in the whole series of volumes, that would mean your subject moved several mm out of the field of view during the EPI run. So the whole brain might have been covered in the first EPI volumes still, but not at the end of the run.
The motion might be correlated with one of the tasks, producing artificial activations. This might already be the case if you have a block design with few repetitions and the patient moved a lot during a single block. At the same time, true activations (as of a much smaller magnitude) would be lost / activation maps might become noisy. Leaving this aside, the more complex the task, the more difficult to detect an interpretable activation pattern on single-subject level. But this can't explain activations in the skull. So I'd say, check the realignment parameters.
|