Providing Novel Views of Visible Human Data for Anatomy Training

Serving Visible Human Data for Anatomy Training

The Pittsburgh Supercomputing Center is collaborating with Dr. Brian Athey and his team at the University of Michigan to provide high speed NGI access to the National Library of Medicine's Visible Human data for use in anatomy classrooms. In this setting, groups of anatomy students will simultaneously refer to different parts of the Visible Human data to help with their studies of actual cadaver anatomy. The Pittsburgh portion of the project provides expertise and assistance in the development and implementation of technologies such as networking, data structures, visualization and high performance computing that are essential to the success of the project.

The Pittsburgh team works closely with anatomists, user interface developers and evaluation teams at the University of Michigan to provide the technology solutions which best fit user needs. The project will provide students and other users with anatomical data views and associated labeling and reference information that has been difficult to access previously. Real time navigation through the data set is an important aspect of the project which drives the need for high performance computing to supply rapid visualization and data retrieval. The project benefits from additional networking projects at PSC including the Web100 research.

The images here show some of the novel viewing capabilities being provided by the project. Along with web based retrieval of Visible Human slice data in original and orthogonal planar formats, the project's network enabled Edgewarp implementation provides arbitrary views constructed directly from the data sets in real time to best illustrate anatomical features.

Collaborating Teams:

University of Michigan: Brian Athey, Victor Wong, Carl Berger, Fred Bookstein, Alex Ade, Walter Meixner, Tom Gest, Deborah Walker, Neil Skov, Bill Green, Terry Weymouth,

PSC: Arthur W. Wetzel, Stuart Pomerantz, Anjana Kar, Jason Sommerfield, Matt Mathis, Nathan Stone

Links:
  • The University of Michigan NGI Visible Human
  • The Visible Human Project®
  • NLM's Visible Human Project: From Data to Knowledge
  • The Stanford Visible Female (SVF)
  • NLM's Next Generation Internet (NGI) Awards
  • NGI Initiative Home Page
  • University of Colorado Center for Human Simulation
  • Networked VH presentation
  • Post-Gazette story
  • Computational Pathology at PSC
  • Power Point presentation to Comp Bio class on April 13

    A curved path following the spinal cord is easily seen from a planar sagittal view.

    Warping the viewing plane to the curved path of spinal cord provides useful information not available directly from planar slice data.

    Without warping only a small segment of the spinal cord can be seen as it intersects any single planar slice. (X1060 Y335 Z950 dy-.3 dz.8)

    A diagonal view of the heart shows its proximity to the spine.

    A synchronized web interface provides access to coordinated orthogonal data planes making it easier to locate features and understand their 3D relationships.

    The Edgwarp-3D software, developed by University of Michigan team members Fred Bookstein and Bill Green, is being extented beyond the manipulation of constrained regions, such as the sholder joint illustrated here, to cover the entire body by dynamically loading new data over the network as a user's region of interest moves.

    Support for PC Windows and Mac OS X platforms is being provided by an Edgewarp

    Screen cap for Jenda

    Screen cap for Jenda