hello,
I've got a error when I process 3dDWItoDT.
I first used the align_epi_anat.py for the registration, and used the output from align_epi_anat.py for 3dDWItoDT.
It worked well, but two of my subjects were failed to do correct registration.
so I decided to do FSL flirt for these two, and I checked that T1 and DTI are registered correctly.
After that, I converted the flirt output file to AFNI format and tried to do 3dDWItoDT; but I got an error.

It says "ERROR: Error - Dataset must have number of sub-briks equal to one more than number. FATAL ERROR: of gradient vectors (b0+bi)!

Since the command worked fine previously (when I registered with align_epi_anat,py), I guess it is nothing to do with bvec file or command line itself. I checked the voxel size and orientation of output file from align_epi_anat,py with output file from FSL flirt using 3dinfo, there is no difference. The only difference I noticed was output from align_epi_anat.py has 65 sub-bricks while output from FSL flirt has only one sub-brick. The storage size is also different.

Could you please help me to solve this situation? you may recommend other way for the registration or help me to understand why FSL flirt output has only one sub-brick.

Thank you for your kindness.


LK