Visualization of Diffusion Tensor Imaging Data
Jorik Blaas, Charl P. Botha, Frits H. Post
Data Visualization Group,
Faculty of Electrical Engineering, Mathematics and Computer Science
Delft University of Technology
Tensor Imaging (DTI) is an MRI-based technique for measuring water
diffusion in living tissue. The diffusion is represented by a tensor at each
voxel in a volume. The main application of this technique is in the imaging of
the white matter of the brain, where water diffuses more rapidly along the
neuronal axons than in the perpendicular direction. This phenomenon is used
to visualize the fibrous structure of the white matter.
We first show a series of techniques for progressive visualisation
of diffusion data: diffusing particles, diffusive spots, and a diffusion
distance metric for quantifying connectivity. These techniques use as much
as possible the full multi-directional diffusion information. Also, the
techniques are efficient and can be used for interactive exploration.
DTI is often used to determine trajectories of fibre bundles, or neuronal
connections between regions, in the brain. However, the resulting visualisations
can be complex and difficult to interpret. An effective approach is to pre-
determine trajectories from a large number of seed positions throughout the
white matter (full brain fibre tracking), and to offer facilities to aid the
user in selecting relevant fibre bundles.
It is crucial for the use and acceptance of this technique in clinical
studies: that the selection of the bundles by brain experts should be
interactive, supported by real-time visualisation of the trajectories
registered with anatomical MRI scans. Also, the fibre selection should be
reproducible, so that different experts will achieve the same results.
We present a practical technique for the interactive selection of fibre-
bundles using multiple convex objects that is an order of magnitude faster
than similar techniques published earlier. We also present the results of a
clinical study with ten subjects that show that our selection approach is
highly reproducible for fractional anisotropy (FA) calculated over the
selected fibre bundles.