Dwi pre-processing, no reverse phase-encoding scans

Hi MrTrix experts,

I have a query relating to the initial pre-processing steps of DWI. Currently I have a dataset containing no reverse phase-encoding scans and solely have nii.gz images in the AP direction. Is it possible for me to continue the processing of these images (e.g. dwidenoise/dwifslpreproc etc…) and still have a reliable and valid analysis?


1 Like


If you have a high resolution anatomical scan, it is possible to use this for the distortion correction using the SynB0-DISCO method. Presently dwifslpreproc does not have the option to directly use those outputs (TOPUP field coefficients), so you will have to call eddy yourself with the appropriate inputs. I believe this will change in an upcoming release.

From my own simulations, this method seems to work better than using a gradient echo fieldmap. I will leave the reliability and validity question to the methods people :slight_smile:

Let me know if you need any more details on how to script this.


Hi Nick,

Thank you so much for your input! It’s been very useful. If I may take you up on the matter, could you please provide some more details on how to script this in bash?

Kind Regards,

DM-ed you!

Dear Mr Trxix experts,
I met the same question, my data only has a b0 value in the AP direction. Could you please provide some more details on how to script the Synb0-DISCO in bash?


I also have similar issue.

I only have single PE direction diffusion MRI data and also I don’t have T1 weighted images. How can I remove distortion? Thank you very much.

I am waiting for your kind response.

Warm Regards,