[Mristudio-users] AIR/MI/theory AIR, which is the best?

[Min Liu]

Dear MRIStudio users,

I have four human brain datasets that are scanned sequentially (back
to back) of a single subject using the same DTI protocol. I want to
get an averaged dataset out of the four. Before doing that,
motion/eddy current correction is definitely needed. I am wondering
among the three methods provided by DTIStudio, i.e. AIR/MI?theory AIR.
By visual inspection, there is rotation/shifting between the very
first dataset and the very last by about 2 voxels. For this subtle
movement, is there a conclusion so far regarding which coregistration
is the best (without introducing errors! Sometimes AIR does more harm
than good by misalignment.) Thank you for your thoughts!!

Min