Information

Abstract

In this thesis a combined visualization of dense clinical data like 3D CTA (Computed Tomography Angiography) combined with co-registered real-time images of medical intervention applications is presented. The main challenge here is to provide a merged visualization that allows sucient spatial perception of the important parts, as derived from the pre-operative data, while not occluding the information in the real-time image embedded within the volume. This work presents a new approach of importance denition for volumetric data and how this importance can be used to create a feature-emphasized visualization. Furthermore the viewpoint and the position of the intervention image is used to generate a contextual cutaway which inuences the density of the visualization to avoid an occlusion of the real-time image by less important parts of the volumetric data.

Additional Files and Images

Additional images and videos:
Importance Control
Importance Control		 Result image (CT)
Result image (CT)		 Result image (MRI)
Result image (MRI)
Video
Video

Additional files:
Masterthesis
Masterthesis




BibTeX

Download BibTeX-Entry
@mastersthesis{haidacher-2007-idr,
  title =      "Importance-Driven Rendering in Interventional Imaging",
  author =     "Martin Haidacher",
  year =       "2007",
  abstract =   "In this thesis a combined visualization of dense clinical
               data like 3D CTA (Computed Tomography Angiography) combined
               with co-registered real-time images of medical intervention
               applications is presented. The main challenge here is to
               provide a merged visualization that allows sucient spatial
               perception of the important parts, as derived from the
               pre-operative data, while not occluding the information in
               the real-time image embedded within the volume. This work
               presents a new approach of importance denition for
               volumetric data and how this importance can be used to
               create a feature-emphasized visualization. Furthermore the
               viewpoint and the position of the intervention image is used
               to generate a contextual cutaway which inuences the density
               of the visualization to avoid an occlusion of the real-time
               image by less important parts of the volumetric data.",
  address =    "Favoritenstrasse 9-11/186, A-1040 Vienna, Austria",
  school =     "Institute of Computer Graphics and Algorithms, Vienna
               University of Technology",
  month =      aug,
  URL =        "http://www.cg.tuwien.ac.at/research/publications/2007/haidacher-2007-idr/",
}