We recently noticed some ghosting in our multi-echo EPI fMRI scans. Anyone have any recommendations for dealing with it on the acquisition and/or preprocessing side of things? Thanks so much in advance!
Can you share some more information about your ME-EPI protocol? Things like TR, TEs, multiband factor, number of slices, etc.
In addition to that information, could you share an image of the effect? Even a single slice showing this would be useful. Does it worsen over scan time? Is it present in the SBRef images (if those exist)?
I can imagine that if the subject moved excessively then the initial shim could no longer be accurate, and that could contribute to ghosting.
Try changing the echo spacing a bit. Sometimes your protocol is hitting close to a scanner resonant frequency and that increases phase errors between lines and hence ghosts.