Diffusion tensor MRI is a new imaging modality with the potential to measure fiber-tract trajectories in fibrous soft tissues such as nerves, muscles, ligaments and tendons. However, the datasets produced are volumetric, with six values at each spatial location, and hence a significant challenge to visualize and understand. We are working on new geometric models that could well relate to the interesting structures in the tissue while representing the DTI data. We are also developing a virtual environment that displays geometric representations of the volume data and are gathering feedback on appropriate interaction and visualization techniques to tailor the prototype to two application areas. The applications we are working toward include studying changes in white-matter structures after gamma-knife capsulotomy and pre-operative planning for tumor surgery.

Several doctors have used our application and from their feedback, we concluded that the geometric models we used to represent the linear anisotropies well correlate with the major neural fiber bundles, the cave helps to visualize the complex geometric models, and traditional 2D section along with the 3D models helps visualizing the data.

Future work includes new geometric models that can get rid of the hard thresholding we now use, new scheme of interactivity in the cave, and tuning our application towards two application areas.

Search the Graphics Group Web Pages