I'm Gabriella, a PhD student in Neuroscience and Imaging.
I am currently working on coregistering anatomical scans with DTI scans (FA and MD maps) using bbregister. However, there are some details in the pipeline that I find unclear.

First, is it correct to use the DTI scan file directly as the --mov input? I came across discussions in the mailing list mentioning an intermediate step involving the lowb volume—specifically, performing bbregister first with the anatomical scans and the lowb volume, and then using mri_vol2vol to apply the registration matrix to the DTI map. I am unsure whether this step is always necessary or if it should only be performed when the direct coregistration (bbregister between the anatomical scan and the DTI map) is inaccurate.

Additionally, regarding the initial registration parameter (--init-coreg), what criteria should be used to decide between the default option and alternatives such as --init-fsl?

Thank you in advance for your time and assistance. I look forward to your response.