Preprocessing and Volume Rendering of 3D Ultrasonic Data

  • Authors:
  • Georgios Sakas;Lars-Arne Schreyer;Marcus Grimm

  • Affiliations:
  • -;-;-

  • Venue:
  • IEEE Computer Graphics and Applications
  • Year:
  • 1995

Quantified Score

Hi-index 0.00

Visualization

Abstract

3D-ultrasound is one of the most useful, noninvasive, non-radiative tomographic techniques. Rendering 3D models from ultrasonic data sets is nontrivial due to the apparently "noisy & blurry" nature of ultrasound. In this study, we investigate ways of creating sharp and smooth appearing surfaces from sets of 3D ultrasonic data. In order to improve image quality, preprocessing filters such as speckle, median, and Gaussian are applied off line. By using semi-automatic segmentation tools, interesting portions of 3D ultrasonic data can be isolated within a short time. The presented volume rendering system, implemented completely in software, is capable of generating high quality images within a few seconds on standard desktop workstations or even PC's, thus making the approach practical for clinical applications.