Dear Carlos,
> Dear FSL experts,
> I have a question about diffusion data acquisition for susceptibility-distortion correction with topup. Do I really need to acquire two datasets with reversed encoding direction? Or would it be enough to acquire two DWI datasets with identical PE (with their corresponding b0 images) plus a set of b0 images with reversed encoding?
in order to calculate a fieldmap to use for correcting for susceptibility distortions you only need to collect a pair of b0 images with reversed PE.
If you also want to use eddy to correct for movement and EC distortions you will need to
1. Acquire all diffusion weighted images with reversed PE
OR
2. Make sure that your diffusion gradients are optimised on the whole sphere.
The latter alternative is typically the more practical. If you don’t know the distinction between optimising on the whole and the half sphere have a look at the online help for eddy.
Jesper
>
> The reason I ask is that the datasets I'm about to acquire are follow-up data for a longitudinal study, so I can't reverse the PE of one entire DWI acquisition. If obtaining just an additional set of b0 with inverse encoding is enough, I could use the data without topup correction for longitudinal analyses, and with topup for cross-sectional studies.
>
> Thanks
>
> Carlos
|