<--back--   home   --next-->

So what are our goals in looking at this kind of data? Basically, we would like to visualize it in a way that shows the three-dimensional structure and shape of the anisotropic regions, especially the white matter fiber tracts. This has a number of useful applications- obviously to improve our understanding of neuroanatomy, but also for the sake of surgical planning and cognitive science. The challenge is that the tensor contains a great deal of information- it has six degrees of freedom. If you have one tensor, you can use an ellipsoid, but what about a full three-dimensional field of them? There's a trade-off between showing enough of the information so you're conveying a good sense of what's going on, versus overloading the information with so much information that its either a visually cluttered opaque block, or is so dense that its unintelligible.