Weird results of performing susceptibility distortion correction on the EPI

Hi all,
I performed fieldmap-based distortion correction in my preprocessing with fmriprep. However, it seems to cut off some of the frontal pole regions. I’m wondering if this is normal.
Thanks a lot for your help!

Hi, it is very hard for me to pick any difference just looking at these two screenshots. It looks like there is little distorsion to correct.
I can’t really answer to you question as I don’t really see an effect of the susceptibility distorsion correction here. Maybe what you see as a cut off is just a regular dropout (loss of signal due to within slice dephasing close to air cavity typical for EPI acquisitions?). Just a guess.
Could you tell us to which slice to look at in particular?

What kind of fieldmaps are you using? fMRIPrep currently has some issues with using EPI (blip up/blip down) fieldmaps. A future release will use topup to handle these, and the next major release (21.0.0+) will feature an overhaul of SDC workflows, but for now it is not implemented in fmriprep.

Hi all,
Thanks a lot for your help!
@jsein please the following screeshots for a better comparison. Briefly, before SDC the frontal pole “extended” to the read line, whereas after SDC the EPI signal “shrank” a little bit.

@Steven I’m using a phasediff and two magnitude images. The PhaseEncodingDirection were all P–>A (i.e., “j” in json files) including the functional data.


I think it is expected that the frontal lobe shrank after SDC correction. Indeed P>A acquisition has a tendency to “inflate” this area, so SDC correction shrank it to its original shape. More importantly, do you feel that the functional image after SDC gets closer to anatomical outline drawn in blue?