ASL Motion Correction

Dear AFNI,

I am working with arterial spin labeling data and each dataset contains a series of interleaved control and labeled volumes (e.g., C,L,C,L,C,L,C,L,C,L,C,L…). Because the mean signal intensity of the control and labeled volumes is different, the signal appears zig-zagged - which makes motion correction difficult. I would like to use 3dvolreg to perform motion correction of the control and labeled volumes separately, using the first volume as the base for both, but ultimately I need to have produced one motion corrected interleaved dataset in original order. I’ve tried using a sub-brik selector with 3dvolreg ‘[0…$(2)]’ and ‘[1…$(2]’ but as you know this method produces a motion corrected output dataset containing only those selected volumes. Do you have any advice or guidance for how I can use AFNI to accomplish this goal?

Any help is greatly appreciated,
Thank you!

Hi Alfonso,

I would except you do not really need a CLCLCL time
series, do you? It would seem more likely that they
would be combined someway after the volreg step.

If that is the case, process the L volumes separate
from the C volumes, performing registration on the
L volumes. Then apply those same parameters to the
C volumes.

How does that seem? If that does not seem correct,
would you please provide more details on how you
plan to use them?

  • rick