Diffusion Tensor MRI (DTI) is a special MR imaging technique where the second order symmetric diffusion tensors whose principal eigenvectors are correlated with the underlying fiberous structure (eg. the nerves in brain), are computed based on Diffusion Weighted MR Images (DWI). The principal of DWI, on the other hand, is mean MR signal suppression due to out-of-phase spins of randomly moving (diffusing) particles in a spin-frequency encoded (diffusion sensitive gradient B-fields) volume. The computed tensors describe the local mean (within a given voxel) diffusion pattern of water molecules via a 3D Gaussian distribution in space. The most common analysis and visualization technique is tractography, which is a numerical integration of the principal diffusion direction (PDD) that attempts to reconstruct fibers as streamlines. Despite its simplicity and ease of interpretation, tractography algorithms suffer from several drawbacks mainly due to ignoring the stochastic nature of the data and using the PDD only. An alternative to tractography is connectivity which aims at computing probabilistic connectivity maps based on the above mentioned 3D Gaussian distribution as described by the DTI data. However, the resulting maps are usually ambiguous and hard to interpret.

The presentation will start with an introduction of DT-MRI data followed by a discussion of connectivity and tractography approaches. Then a physical model for DTI connectivity analysis, namely the Lattice-of-Springs (LoS) model, will be introduced. The Split & Merge Tractography (SMT) method, an attempt to find a compromise between the connectivity and tractography approaches, will be presented next. We will conclude with remarks about some of the other peculiarities of DTI processing.