Hi Christian –
That’s great. One thing to make sure is that the order of the volumes is the same as is in your bvec/bval files. I’m not sure how best to do that. It might be safer to re-run dcm2nii on your original DICOM files, and make sure it outputs a 4D volume rather than individual 3D volumes. Are you using the latest version of dcm2nii?
We do have our own programs for merging DWI files (see here: http://neuroimage.usc.edu/neuro/Resources/BDPAddons#mergeDWIs), which is useful when a set of diffusion data is acquired as two different acquisitions. However, this program expects to have one bvec and one bval file per nii file. I think this would be different from how your data currently are stored.