 |
 |
Dear fsl experts, I do have some questions about the bedpostX workflow especially about the different models and how to work with them. From the DTI FDT (BedpostX) documentary I know that there are two models, first one for the normal approach and the second one for multishell data. Here in the mailing list (and in paper) I have read different things about the models and I also have read that there is a third model. But I can't find some more information on the fsl webpage about it. So here are my questions: 1.How many models are there? 2.What exactly are the models doing, which technique(algorithm) do they use (in general)? (For me its not really clear from the technique report which is linked on the bedpost documentary) 3.What and where is the difference about/between parametric deconvolution (ball and stick) and non-parametric deconvolution (SD) method? (I have read Behrens et al, NeuroImage 2007. and Jbabdi et al. MRM 2012 but I still do not understand it well) 4.How does a input file for a multishell approach look like? I do have 2 preprocessed volumes (different b-values) with each 60 dwi + 7 non dwi. Do I just have to merge the volumes and their bvec, bvalues files, then title them as data.nii etc. and in the end start the bedpostX command with model=2 ? 5.What happens when I do the same but with model=1? Is it not also a multishell approach? Thanks for any help! Best Max