I'm a tiny bit confused about what comes out of dcm2nii when applied to a DTI scan. I get a nifti with my data, and I get a bvec file, but the values in that file are pretty different from the values I see in the header for DiffusionGradientOrientation (0018, 9089) tag, which I can extract via software like MATLAB or PMod fairly easily.

Maybe this is a scanner / patient coordinate issue or maybe I'm just looking at the wrong tag? Using what comes out of dcm2nii gives substantially worse tractography in software such as trackvis, for example.

Anyway, any help here would be greatly appreciated!