A fuzzy connectivity tree for hierarchical extraction of venous structures

Authors:
Silvana Dellepiane;Lorena Novelli;Michele Bruzzo;Marco Antonelli
Affiliations:
University of Genoa, Department of Biophysical and Electronic Engineering, Genova, Italy;University of Genoa, Department of Biophysical and Electronic Engineering, Genova, Italy;University of Genoa, Department of Biophysical and Electronic Engineering, Genova, Italy;University of Genoa, Department of Biophysical and Electronic Engineering, Genova, Italy
Venue:
SCIA'03 Proceedings of the 13th Scandinavian conference on Image analysis
Year:
2003

Citing 3
Cited 0

Extraction of intensity connectedness for image processing

Pattern Recognition Letters
Fuzzy connectedness and object definition: theory, algorithms, and applications in image segmentation

Graphical Models and Image Processing
Nonlinear image labeling for multivalued segmentation

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In Magnetic Resonance Angiography (MRA), blood vessels show very high grey levels resulting from the use of a contrast agent. This leads to develop a formula, named β-connectedness, derived from fuzzy-connectedness theory [1],[2]. β-connectedness is described here as a specific case of more general χ-connectedness [3], for it allows one to improve the tracking of bright structures and, consequently, vessel extraction. In the computation of fuzzy connectedness, a widely used approach is based on an adaptive growing mechanism that follows the best paths starting from a reference seed point. As a consequence, a hierarchical tree is generated. Contrary to the aforesaid approach, here we propose to exploit such a tree information and we define the connectivity level as an additional parameter useful in the analysis of image data. The method presented in the paper deals with a new kind of connectivity that depends on the position of each pixel in the growing tree. The detection of fine structures is thus improved, as demonstrated by preliminary results on 3D MRA volumes.