Here is the answer from TORTOISE folks if someone else needs:
Quote
The following instruction to run DRBUDDI is with the understanding that the PA volume is a b0. 1) Please open the PA b0 and create copy of that volume. 2) create a 4dnifti file from these two volumes. Create bvecs and bvals file for this which is essentially consists of zeroes (bvals row of 2 zeroes and bvecs will be 3x2 (RxC) zeroes). 3) import this PA b0.nii using import nifti in DIFFPREP. This will create a proc folder that consists of .list, .nii and .bmtxt for PA. 4) run diffprep on the AP data (you may run with to be drbuddied tag 1, to not write out the final steps). PA data can remain as is without any DIFFPREP processing. 5) now in DRBUDDI use AP_proc.list, PA.list and provide a structural image. Please also use the following tags: --n_dwis 0 and --final_signal_modulation LS 6) In DRBUDDI_proc the corrected AP and PA will be separated and not combined but the AP files would have used the information from PA b0 to correct for EPI distortions. Ideally it is good to have more bvalues in the blip down that can match the blip up but with one, some amount of correction is achievable. please check the results i.e AP_b0_corrected_final.nii with respect to the structural_used.nii file. You will have to use the AP_DRBUDDI corrected final.list file to compute the tensors.
Thank you
2086