FSPGR

[stshea]

Gibbs ringing is normally caused by truncation of data before the Fourier transform. From a MR protocol perspective, this results from low spatial resolution (i.e. large voxels) and/or zero-filling interpolation methods that do not smooth the transition from k-space regions that have signal to the zero-filled regions that do not.

You did not list the number of partitions or slices that you were acquiring. If it's a very small number (10 or less), it's possible that this is the cause of your Gibbs ringing. In addition, many scanner manufacturers' software automatically interpolates in the "z" or slice direction for 3D scans. So, you may be thinking you are acquiring 16 slices at 0.8 mm resolution but you are actually only acquiring 8 slices at 1.6 mm. Adding partial fourier accelerations can further complicate this. I would suggest increasing the number of slices/partitions that you are acquiring and possibly consider increasing your through-plane resolution to see if that solves the problem.