Hybrid framework for medical image segmentation

Authors:
Chunyan Jiang;Xinhua Zhang;Christoph Meinel
Affiliations:
Hasso-Plattner-Institut, University of Potsdam, Potsdam, Germany;LIASIT, Luxemburg;Hasso-Plattner-Institut, University of Potsdam, Potsdam, Germany
Venue:
CAIP'05 Proceedings of the 11th international conference on Computer Analysis of Images and Patterns
Year:
2005

Citing 6
Cited 0

Computational geometry: an introduction

Computational geometry: an introduction
Fronts propagating with curvature-dependent speed: algorithms based on Hamilton-Jacobi formulations

Journal of Computational Physics
Fuzzy connectedness and object definition: theory, algorithms, and applications in image segmentation

Graphical Models and Image Processing
Using Prior Shapes in Geometric Active Contours in a Variational Framework

International Journal of Computer Vision
Boundary Finding with Correspondence Using Statistical Shape Models

CVPR '98 Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Segmentation of biological volume datasets using a level-set framework

VG'01 Proceedings of the 2001 Eurographics conference on Volume Graphics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Medical image segmentation is essential step for many image processing applications. In this paper, we present a hybrid framework designed for automated segmentation of radiological image, to get the organ or interested area from the image. This approach integrates region-based method and boundary-based method. Such integration reduces the drawbacks of both methods and enlarges the advantages of them. Firstly, we use fuzzy connectedness method to get an initial segmentation result and homogeneity classifier. Then we use Voronoi Diagram-based to refine the last step's result. Finally we use level set method to handle some vague or missed boundary, and get smooth and accurate segmentation. This hybrid approach is automated, since the whole segmentation procedure doesn't need much manual intervention, except the initial seed position selection for fuzzy connectedness segmentation.