Dear Noelia,
>
> I have read the technical note for probabilistic tractography (Behrens et al., 2003) and was hoping that someone could help me to have a better conceptual understanding. For example, I would be grateful if someone could explain me whether BEDPOSTX is using a partial volume model to calculate probability density functions (pdfs) and then using and Markov Chain Monte Carlo sampling to build up distributions of voxel-wise principal diffusion directions (theta, phi)
it uses MCMC to build distributions of theta and phi (and of the other model parameters). There isn’t a partial volume model, and the pdfs are the distributions. For more recent versions of bedpost there are more than one distribution for theta and phi, corresponding to the multiple fibres.
> which are later sampled in PROBTRACKx to generate probabilistic streamlines. So that the outcome is a connectivity distribution (for a given voxel, number of samples that pass between the seed voxel and that voxel).
Yes, and at this stage you could say that there is a partial model volume, insomuch that when a step takes you to anything else than a voxel centre (which almost every step will) there is a probabilistic selection of which of the surrounding voxels from which to sample the distributions of theta and phi.
>
> If that is correct, would it make sense to calculate microstructural properties (FA, MD, AD and RD) which are derived by fitting a diffusion tensor model at each voxel within the connectivity distributions obtained with probabilistic tractography?
>
> This is what I would do for seeds that are in deep gray matter since as the technical note points out these tend to have low anisotropy.
I don’t quite undertand these two questions?
Jesper
>
> Thank you in advance for your clarification.
>
> Best,
> Noelia
>
> ########################################################################
>
> To unsubscribe from the FSL list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|