Toggle Main Menu Toggle Search

Open Access padlockePrints

Local phase approaches to extract biomedical networks

Lookup NU author(s): Professor Boguslaw ObaraORCiD


Full text for this publication is not currently held within this repository. Alternative links are provided below where available.


Many biomedical applications require detection of curvilinear networks in images, and would benefit from automatic or semiautomatic segmentation to allow high-throughput measurements. Here we discuss a contrast independent approach to identify curvilinear structures based on oriented phase congruency, the Phase Congruency Tensor. We show that the proposed approach is largely insensitive to intensity variations along the curve, and provides successful detection within noisy regions. Moreover, we demonstrate that the proposed approach may be used in a wide range of curvilinear and non-curvilinear feature enhancement and detection methods, particularly where tensor representation of the image is explored. The performance of the Phase Congruency Tensor-based methods is evaluated by comparing it with state-of-the-art intensity-based methods on both synthetic and real images of biomedical networks. © 2012 IEEE.

Publication metadata

Author(s): Obara B, Fricker M, Grau V

Publication type: Conference Proceedings (inc. Abstract)

Publication status: Published

Conference Name: 9th IEEE International Symposium on Biomedical Imaging (ISBI 2012)

Year of Conference: 2012

Pages: 1796-1799

Online publication date: 12/07/2012

ISSN: 1945-8452

Publisher: IEEE


DOI: 10.1109/ISBI.2012.6235931

Library holdings: Search Newcastle University Library for this item

ISBN: 9781457718588