[Mrtrix-discussion] MRtrix / Camino interoperability

Gavin Winston g.winston at ucl.ac.uk
Thu Jun 24 04:18:30 PDT 2010


Previously, I have been using the multitensor model and probabilistic 
tractography (PICo) in Camino. I would now like to use the CSD code from 
MRtrix, but ideally still perform the tractography within Camino (so I can 
compare the results more easily).

It is possible to save the output of csdeconv as a 4D Analyze file. Camino 
appears to include a utility shformatconverter that should be able to 
convert these coefficients into Camino order. However, it is not working as 
expected. Does anybody have experience of doing this? If it is more 
appropriate to email the Camino list, I will do so.

The steps I am taking are running shredder to convert to Big endian (as I 
run MRtrix on a Windows box), scanner2voxel to convert the coefficients so 
all the coefficients for each voxel are together, shformatconverter to 
reorder the coefficients, then sfpeaks. However, the result is nonsense.

I assume that the order of the coefficients is Y(0,0); Im(Y2,2); Im(Y2,1) 
etc as documented by csdeconv -h as this is what shformatconverter seems to 
expect. However, the documentation manual within the package gives a 
completely different order.

Gavin 



More information about the Mrtrix-discussion mailing list