Fused multi-volume DVR using binary space partitioning

  • Authors:
  • Stefan Lindholm;Patric Ljung;Markus Hadwiger;Anders Ynnerman

  • Affiliations:
  • Linköping University, Sweden and Siemens Corporate Research;Siemens Corporate Research;VRVis Research Center, Austria;Linköping University, Sweden

  • Venue:
  • EuroVis'09 Proceedings of the 11th Eurographics / IEEE - VGTC conference on Visualization
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Multiple-volume visualization is a growing field in medical imaging providing simultaneous exploration of volumes acquired from varying modalities. However, high complexity results in an increased strain on performance compared to single volume rendering as scenes may consist of volumes with arbitrary orientations and rendering is performed with varying sample densities. Expensive image order techniques such as depth peeling have previously been used to perform the necessary calculations. In this work we present a view-independent region based scene description for multi-volume pipelines. Using Binary Space Partitioning we are able to create a simple interface providing all required information for advanced multi-volume renderings while introducing a minimal overhead for scenes with few volumes. The modularity of our solution is demonstrated by the use of visual development and performance is documented with benchmarks and real-time simulations.