 |
 |
Hi > > I would like to know if there are limitations in using the above methods with regards to voxel size. I have data with voxel dimensions of 2mm x 2mm x 5mm. I presume smaller and isometric would be better, but can I use this dataset? In my opinion, very anisotropic voxels is a very bad idea. In probabilistic tractography, for example, it would imply that the uncertainty on tract orientations will depend on the position of the head in the scanner... > If one has groups of subjects to compare, is there some metric in the output to do this on? How about the connectivity distribution between seed and target sites in each subject? Generally, people compare DTI outputs (FA, MD, eigenvalues). Using connectivity distribution is also an option, but probably not an end in itself, rather a starting point for investigating what is driving the differences you observe. This is because these connection probabilities do not relate (at least not obviously) to tract "strength", but rather tell you how likely you are to have a connection by chance. Several factors could be driving this likelihood, and you probably want to investigate some of those factors in order to understand what is driving group differences (factors such as micro-structure, uncertainty, seed/target size, etc.) Cheers, Saad. > > Thanks.......J > -- Saad Jbabdi University of Oxford, FMRIB Centre JR Hospital, Headington, OX3 9DU, UK (+44)1865-222466 (fax 717) www.fmrib.ox.ac.uk/~saad