DTI metrics derived from multi-shell data with high b-values

[MRTanja]

• What is the current behaviour?
  In 02_proc-dwi.sh it seems like DTI metrics are calculated regardless of the nature of the data, however for multi-shell data the higher shells above b=1500 should not be used for tensor fitting because the assumption of Gaussian diffusion does not hold, so I was wondering if I'm misunderstanding something or if this is an issue? If this is indeed how the script works then I would implement something for my dataset whereby DTI metrics would be derived from a subset of the data including only the lower shells, but I wanted to make sure I understood the script correctly first.

• What is the motivation / use case for changing the behaviour?
  Literature suggests that for multi-shell data the higher shells above b=1500 should not be used for tensor fitting because the assumption of Gaussian diffusion does not hold. Therefore when multi-shell data is processed with 02_proc-dwi.sh, only lower shells (e.g. below b=1500) should be selected.