Hi Rob
Thanks for your detailed response.
In this particular scenario, there’s some chance that it may be preferable to not use the b=0 volume; but this is entirely speculative on my part. In such a scenario, using something like dwi2response tournier is more convenient, because the derived response function contains on the coefficients for the b>0 data,
Sorry i forgot to read this paragraph accurately. Since my data have a solitary b0 volume ( one b0 volume and 64 volume with b=1000), I’ve decided to use the tournier algorithm for creating the response function in order to avoid the possible error in estimating the response function.
If you wish to discard the b=0 data and utilise only the b=1000 data, in a single-tissue deconvolution, then yes, that should work. But as I said, I’m not strictly advocating for that approach; I’m merely pointing out that in the instance where one has only one b=0 image rather than several, it’s not strictly guaranteed that a two-shell two-tissue deconvolution will be preferable to a single-shell single-tissue deconvolution.
Considering my dwi data, I use this approach for dwi2fod msmt to gain the benefit of this algorithm compare with csd as you mentioned previously (Fixel-based analysis results)