Inferring vascular structures in coronary artery X-ray angiograms based on multi-feature fuzzy recognition algorithm

Shoujun Zhou*, Wufan Chen, Jiangui Zhang, Yongtian Wang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

The multi-feature fuzzy recognition (MFFR) algorithm was presented to infer the vessel structures, in the context of X-Ray Angiograms (XRA) of the coronary artery. In the modeling, a multi-feature metrics (MFM) was firstly established to describe the local configuration; then the membership degree of MFM-based fuzzy subsets was defined, and the fuzzy recognition operator was constructed. The MFFR algorithm can correctly infer four kinds of vessel structures including vascular ends, segments, bifurcations and crossovers. The results are satisfying: on average 91.1% of the testing vessel lengths in medium quality images are automatically delineated as well as their structures being correctly inferred with point-wise.

Original languageEnglish
Title of host publicationMedical Imaging and Augmented Reality - Third International Workshop
PublisherSpringer Verlag
Pages325-332
Number of pages8
ISBN (Print)3540372202, 9783540372202
DOIs
Publication statusPublished - 2006
Event3rd International Workshop on Medical Imaging and Augmented Reality - Shanghai, China
Duration: 17 Aug 200618 Aug 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4091 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference3rd International Workshop on Medical Imaging and Augmented Reality
Country/TerritoryChina
CityShanghai
Period17/08/0618/08/06

Keywords

  • Coronary artery X-ray angiograms
  • Multi-feature fuzzy recognition
  • Vessel structure inference

Fingerprint

Dive into the research topics of 'Inferring vascular structures in coronary artery X-ray angiograms based on multi-feature fuzzy recognition algorithm'. Together they form a unique fingerprint.

Cite this