Three-Dimensional Coronary Artery Centerline Extraction and Cross Sectional Lumen Quantification from CT Angiography Images

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

Abstract

Automatic centerline extraction based on 3D coronary artery segmentation results is a very important step before quantitative evaluation of intravascular lumen cross-section. In this paper, a method based on the combination of fast marching and gradient vector flow (GVF) is proposed to extract the centerline of the complete coronary artery tree in 3D angiographic images. With the centerline of blood vessel, we propose an automatic method to extract the cross-section of blood vessel lumen. This method calculates the tangent vector based on the two adjacent centerline points before and after the midline point, and then calculates the cross-sectional equation through the centerline point, and then obtains the cross-sectional contour of the cross-section and the surface mesh of blood vessel. The new method is designed to extract the cross-section of 3D intravascular lumen in real physical coordinates, which avoids the traditional interpolation processing in pixel coordinates and improves the accuracy of cross-section extraction. Given the accuracy and efficiency, the proposed coronary artery lumen area measurement algorithm can facilitate quantitative assessment of the anatomic severity of coronary stenosis.

Original languageEnglish
Title of host publicationIntelligence Science and Big Data Engineering. Visual Data Engineering - 9th International Conference, IScIDE 2019, Proceedings, Part 1
EditorsZhen Cui, Jinshan Pan, Shanshan Zhang, Liang Xiao, Jian Yang
PublisherSpringer
Pages238-248
Number of pages11
ISBN (Print)9783030361884
DOIs
StatePublished - 2019
Event9th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2019 - Nanjing, China
Duration: 17 Oct 201920 Oct 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11935 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference9th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2019
Country/TerritoryChina
CityNanjing
Period17/10/1920/10/19

Keywords

  • Centerline
  • Computer-aided diagnosis
  • Coronary artery segmentation
  • Cross sectional lumen area
  • Medical imaging

Fingerprint

Dive into the research topics of 'Three-Dimensional Coronary Artery Centerline Extraction and Cross Sectional Lumen Quantification from CT Angiography Images'. Together they form a unique fingerprint.

Cite this