Warping artifacts due to PE direction

Hi all!

I’m working with a dataset with AP as the main phase encoding direction, and I also have the reverse PE acquisition for some b0 volumes. The thing is that the AP dwi present the typical warping artefacts and the very frontal region of the brain is missed:

Caption: LEFT: mean b0 with AP. RIGHT: T1w image and the mean b0 overlaid (red).

During the preprocessing steps I executed:

dwifslpreproc dwiPA_denoise_unring.mif dwi_denoise_unring_preproc.mif -pe_dir PA -rpe_pair -se_epi b0_pair.mif -eddy_options "--slm=linear --data_is_shelled"

The output looks the same.

Is there any way to correct this artefact? Any thoughts?

One of the consequences is that, for example, the frontal pole area of the Desikan-Killiany atlas is missed in the dwi.

Thank you!


Hi @stellamarissanchez89,

There seems to be a mismatch:

Sounds like your pe_dir should be AP

-pe_dir PE Manually specify the phase encoding direction of the input series


Hi Steven,

Thanks for your reply. Actually, it was my mistake when I copied the command, I executed it with pe_dir AP.

Do you use the same command? And do you usually get good quality results?

Thanks for your time :slight_smile:


Hi @stellamarissanchez89,

I use qsiprep to preprocess my data, which has generally done a good job of unwarping (although like mrtrix, it also internally calls upon FSL commands).


1 Like