The Community for Technology Leaders
Visualization Symposium, IEEE Pacific (2009)
Beijing, China
Apr. 20, 2009 to Apr. 23, 2009
ISBN: 978-1-4244-4404-5
pp: 145-151
Song Zhang , Mississippi State University, USA
Stephen Correia , Brown University, USA
David F. Tate , Brigham and Women's Hospital, USA
Wei Chen , State Key Lab of CAD&CG, ZJU, China
Diffusion tensor fields reveal the underlying anatomical structures in biological tissues such as neural fibers in the brain. Most current methods for visualizing the diffusion tensor field can be categorized into two classes: integral curves and glyphs. Integral curves are continuous and represent the underlying fiber structures, but are prone to integration error and loss of local information. Glyphs are useful for representing local tensor information, but do not convey the connectivity in the anatomical structures well. We introduce a simple yet effective visualization technique that extends the streamball method in flow visualization to tensor ellipsoids. Each tensor ellipsoid represents a local tensor, and either blends with neighboring tensors or breaks away from them depending on their orientations and anisotropies. The resulting visualization shows the connectivity information in the underlying anatomy while characterizing the local tenors in detail. By interactively changing an iso-value parameter, users can examine the diffusion tensor field in the entire spectrum between the continuous integral curves and the discrete glyphs. Expert evaluation indicates that this method conveys very useful visual information about local anisotropy in white matter fibers. Such information was previously unavailable in tractography models. Our method provides a visual tool for assessing variability in DTI fiber tract integrity and its relation to function.
Song Zhang, Stephen Correia, David F. Tate, Wei Chen, "Visualizing diffusion tensor imaging data with merging ellipsoids", Visualization Symposium, IEEE Pacific, vol. 00, no. , pp. 145-151, 2009, doi:10.1109/PACIFICVIS.2009.4906849
97 ms
(Ver 3.3 (11022016))