[Camino-users] Bayesian tractography interpretation

[Philip A Cook]

You are correct, you get 1 streamline per seed point per iteration, with the principal directions drawn from the posterior distribution once per iteration.

The easiest way to combine these into an image is to pass the tracts to procstreamlines with the -outputacm option. This will give you a total count of the streamlines intersecting each voxel. It's not strictly a probability in that it's not an image in the range [0, 1] that reflects our degree of belief that the seed region is connected to a particular voxel. Qualitatively, it highlights more likely connections and gives a visual representation of the uncertainty introduced by random measurement errors in the voxel principal directions.



On Mar 17, 2014, at 3:24 PM, Mojdeh Zamyadi <mojdehzm at gmail.com> wrote:

> Hi all,
> 
> I have a question about how to interpret the Bayesian tractography results and I apologize if it's naive! If I use a certain number of iterations (eg. 1000), does this mean that there would be 1000 tracts starting from each voxel in the seed region by drawn a random direction each time according to the posterior distribution? ie. would I have 1000 tracts initiating from each seed voxel?!
> 
> And the next question is how can get a probability map of the tracts (I'm assuming "track" command with -bayesian option gives me number of tracts, correct?)
> 
> Thanks a lot,
> -Mojdeh 
> 
> _______________________________________________
> Camino-users mailing list
> Camino-users at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/camino-users