have you looked at all the images to assure that none have any artifacts? Have you tried a subset to see if that will work? All the images have the same geometry, I presume?
doug
Sebastian Moeller wrote:
Dear All,
I encounter some problems with mri_motion_correct2. I scanned 10 structurals of a monkey that was not moved or moving during the whole scan time. Now I expect mri_motion_correct2 to produce nice averages of my ten runs, but it insists upon displacing/rotating some of my scans. That is in one "beautiful" example a single run's volume was turned by 90 degrees. Are there any expert options for mri_motion_correct2 to clue it in on the expected scan quality? A before I forget centos4_x86_64-dev20060224 is the version I am trying.
Ahoi Sebastian