Dear FSL experts,
Based on the FSL wiki and a few emails from Stam in the email archive, it seems it's acceptable for topup to acquire a full data set (e.g., 60 directions with 6 b0 in a particular phase-encode direction) and then acquire a few b0s in the reverse phase encode. An alternative way would be to acquire two "symmetric" 30 direction datasets with reverse phase-encodes.
Looking at the big picture, from geometric distortion correction, to modeling the diffusion parameters in bedpost, to building fiber tract representations in probtrack….is there any suggested acquisition? As usual, we're limited by time and need to select the most efficient/effective method.
My sense is that more unique DWIs are desirable on the whole (as long as there is not a substantial penalty/bias in the distortion correction).
Any advice is much appreciated!
All the best,
Ryan Muetzel
|