Three-dimensional fusion of blood vessel network using automatic registration of shared bifurcation points in multiple ultrasound volumes

Tuan Hung Phan, Shinya Onogi, Kohji Masuda

Research output: Contribution to journalArticle

Abstract

Nowadays, we can obtain 3D vascular network information inside a body using a 2D matrix array ultrasound probe. This information is useful for diagnosis or intra-operation. However, Doppler volume is incomplete because obtained volume dimensions are still limited and Doppler signal strength depends on flow direction. Therefore, to obtain complete vascular network, we propose a method for fusion of multiple 3D Doppler volumes using automatic registration of shared bifurcation points. To evaluate the method, 24 Doppler volumes of four human subjects were acquired. We confirmed that the method can automatically fuse Doppler volumes with 0.53 mm in root mean square error.

Original languageEnglish
Pages (from-to)O-279-O-280
JournalTransactions of Japanese Society for Medical and Biological Engineering
Volume52
DOIs
Publication statusPublished - Aug 17 2014

Fingerprint

Blood vessels
Electric fuses
Mean square error
Fusion reactions
Ultrasonics

All Science Journal Classification (ASJC) codes

  • Biomedical Engineering

Cite this

Three-dimensional fusion of blood vessel network using automatic registration of shared bifurcation points in multiple ultrasound volumes. / Phan, Tuan Hung; Onogi, Shinya; Masuda, Kohji.

In: Transactions of Japanese Society for Medical and Biological Engineering, Vol. 52, 17.08.2014, p. O-279-O-280.

Research output: Contribution to journalArticle

@article{4609ea2e1b7048e6b10134cc6ba064ca,
title = "Three-dimensional fusion of blood vessel network using automatic registration of shared bifurcation points in multiple ultrasound volumes",
abstract = "Nowadays, we can obtain 3D vascular network information inside a body using a 2D matrix array ultrasound probe. This information is useful for diagnosis or intra-operation. However, Doppler volume is incomplete because obtained volume dimensions are still limited and Doppler signal strength depends on flow direction. Therefore, to obtain complete vascular network, we propose a method for fusion of multiple 3D Doppler volumes using automatic registration of shared bifurcation points. To evaluate the method, 24 Doppler volumes of four human subjects were acquired. We confirmed that the method can automatically fuse Doppler volumes with 0.53 mm in root mean square error.",
author = "Phan, {Tuan Hung} and Shinya Onogi and Kohji Masuda",
year = "2014",
month = "8",
day = "17",
doi = "10.11239/jsmbe.52.O-279",
language = "English",
volume = "52",
pages = "O--279--O--280",
journal = "BME = Bio medical engineering / henshu, Nihon ME Gakkai",
issn = "1347-443X",
publisher = "Japanese Society for Medical and Biological Engineering",

}

TY - JOUR

T1 - Three-dimensional fusion of blood vessel network using automatic registration of shared bifurcation points in multiple ultrasound volumes

AU - Phan, Tuan Hung

AU - Onogi, Shinya

AU - Masuda, Kohji

PY - 2014/8/17

Y1 - 2014/8/17

N2 - Nowadays, we can obtain 3D vascular network information inside a body using a 2D matrix array ultrasound probe. This information is useful for diagnosis or intra-operation. However, Doppler volume is incomplete because obtained volume dimensions are still limited and Doppler signal strength depends on flow direction. Therefore, to obtain complete vascular network, we propose a method for fusion of multiple 3D Doppler volumes using automatic registration of shared bifurcation points. To evaluate the method, 24 Doppler volumes of four human subjects were acquired. We confirmed that the method can automatically fuse Doppler volumes with 0.53 mm in root mean square error.

AB - Nowadays, we can obtain 3D vascular network information inside a body using a 2D matrix array ultrasound probe. This information is useful for diagnosis or intra-operation. However, Doppler volume is incomplete because obtained volume dimensions are still limited and Doppler signal strength depends on flow direction. Therefore, to obtain complete vascular network, we propose a method for fusion of multiple 3D Doppler volumes using automatic registration of shared bifurcation points. To evaluate the method, 24 Doppler volumes of four human subjects were acquired. We confirmed that the method can automatically fuse Doppler volumes with 0.53 mm in root mean square error.

UR - http://www.scopus.com/inward/record.url?scp=84939441507&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84939441507&partnerID=8YFLogxK

U2 - 10.11239/jsmbe.52.O-279

DO - 10.11239/jsmbe.52.O-279

M3 - Article

VL - 52

SP - O-279-O-280

JO - BME = Bio medical engineering / henshu, Nihon ME Gakkai

JF - BME = Bio medical engineering / henshu, Nihon ME Gakkai

SN - 1347-443X

ER -