The proposed Biomedical Tele-Visualization research aims to explore an innovative way of performing biomedical visualization. The overall objective of the research is to develop new algorithms and technology so that 3D biomedical visualization can be performed conveniently, interactively, and collaboratively on off-the-shelf desktop computers. The basic ideas of the research are: 1) the separation between the 3D rendering and the 2D display of volume visualization, 2) the employment of high-performance networks for visualization transmission, and 3) the support of realtime interaction and collaboration. More specifically, the operations for visualization computation and rendering are performed on a high-performance server with dedicated graphics hardware. Results of the operations, in the form of image streams, are transmitted to a group of clients over high-performance networks in realtime. The tasks on the client machines are kept minimum, and can be performed by the off-the-shelf desktop computers. Client users can control the visualization operations interactively, collaboratively, and in realtime.

This research is clearly multi-disciplinary, requiring a combined effort of several areas in computer science, in addition to an active collaboration with biomedical researchers. Broadly speaking, the research objective will be achieved by exploring and integrating advanced technologies from the following three areas, namely 3D visualization of large volume-image datasets, realtime multimedia networking, and collaboration management. Recent advances in 3D biomedical data acquisition techniques, high-performance graphics hardware, and high-performance networking have made the proposed approach feasible. Specifically, the research will employ the newly launched Abilene network for realtime delivery of high-bandwidth visualization image streams to geographically scattered users.

Our approach is fundamentally different from existing techniques for interactive visualization of biomedical datasets. Current visualization techniques require direct access to dedicated high-performance graphics hardware. Such a constraint severely limits the far-reaching impact of visualization on a wide range of biomedical applications. The success of the proposed research will significantly advance the accessibility and usability of biomedical visualization, which in turn would have a broad and profound impact on biomedical research and health care.