Yep, that’s one option; however, there’s a number of challenges and limitations with that approach (dwiintensitynorm) to intensity normalisation. Take a look around the forum for some relevant discussions. In any case, having single-shell data is not a show-stopper on this front though: you can perform…
- 2-tissue (WM-CSF) CSD with MSMT-CSD
- 3-tissue (WM-GM-CSF) CSD with SS3T-CSD
…on your data. Both are technically compatible with following up with mtnormalise for intensity normalisation and bias field correction.
That said, there’s a few more complex limitations as well as strong considerations as to final interpretation if you use specifically 2-tissue CSD + mtnormalise though. There are scenarios where one might argue it’s not exactly the right thing to do, and it has interesting implications for the outcome; both in terms of global as well as spatially varying intensity normalisation. This is mostly of concern at lower b-values. This is not really the best place (and timing) to elaborate on such though.