As part of the Visible Human Project , the U.S. National Library of Medicine has recently made digitized datasets of male and female human cadavers available for research and education.
The Mathematics and Computer Science Division ( MCS ) of Argonne National Laboratory and the Academic Computing Services department of the University of Chicago ( UC ) are exploring the visualization of these datasets in networked virtual environments. The software components being used are the "VRen" volume rendering C++ class, developed at UC and MCS and the Nexus-based CAVEComm library for networking virtual environments. The virtual environments being used are the CAVE(TM) and ImmersaDesk(TM) at Argonne. Both environments are driven by the SGI ONYX - the CAVE by a rack-mounted model with 4 MB of texture memory, and the ImmersaDesk by a deskside model with 16 MB of texture memory. VRen uses the 3d texture mapping hardware of the ONYX to volume render the data.
Below are some images produced from the fresh CT scan data of the visible female dataset. They show various structures of the human body made visible by changing the visualization parameters of VRen. The original, fresh CT scan dataset consists of 1734 slices of 512 × 512 2-byte pixels. To display the entire body near-real-time on the deskside ONYX, the dataset was subsampled, with filtering, to 1024 slices of 128 × 128 2-byte pixels. (The application runs somewhat slower on the rack-mounted ONYX.)