Share Email Print

Proceedings Paper

3D segmentation of medical images for computer-aided design and rapid prototyping of orthopedic devices
Author(s): Justin R. Kidder; Bartholomew O. Nnaji
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

The segmentation of medial images is the first step along the series of events that need to take place during the automated design and rapid prototyping of orthopedic devices. The segmentation algorithm presented in this work, the discrete surface model, is an automated 3D segmentation method that attempts to optimize an energy function based on three distinct potentials: a deformation potential that expands the model, an image potential that causes it to stop at image boundaries, and a topology potential that maintains a smooth curvature. Since the deformation of the discrete surface model is local in nature, a global optimization routine encapsulates the deformation of the model using simulated annealing to avoid local energy minima such as are caused by false edges. In addition, the model has no inherent way of checking for self intersections. Global topology is also maintained using a ray-shooting technique that prevents self intersections for occurring in concave models. This segmentation algorithm is ideal for the creation of physical 3D models using rapid prototyping.

Paper Details

Date Published: 2 January 1998
PDF: 12 pages
Proc. SPIE 3201, Sensors and Controls for Advanced Manufacturing, (2 January 1998); doi: 10.1117/12.298004
Show Author Affiliations
Justin R. Kidder, Univ. of Pittsburgh (United States)
Bartholomew O. Nnaji, Univ. of Pittsburgh (United States)

Published in SPIE Proceedings Vol. 3201:
Sensors and Controls for Advanced Manufacturing
Bartholomew O. Nnaji; Anbo Wang, Editor(s)

© SPIE. Terms of Use
Back to Top