Reconstruction and auto-correction of artificial capillary with flow directions from ultrasound volume data

Antoine Bossard, Yuki Sugano, Shinya Onogi, Toshikazu Kato, Nobuhiko Shigehara, Takashi Mochizuki, Kohji Masuda

    Research output: Contribution to journalArticlepeer-review

    Abstract

    Recently, we described a reconstruction method of the blood vessel network (BVN) by 3D thinning to detect vessels bifurcations, which is applied to the control of microbubbles in vivo. However, that method did not include blood flow directions and was only verified on a very simply shaped artificial blood vessel. In this paper we propose a much improved system, this time including blood flow directions and, importantly, assessing the reconstruction correctness by introducing an abstraction method for the BVN. Such a model is then analyzed through graph theory and error patterns for auto-correction of the reconstructed BVN. Data was acquired by a Philips xMATRIX iU22 3D-capable echography and a matrix array probe (X6-1). Flow directions were obtained by using colour Doppler information. Additionally, we conducted several validating experiments: first, we performed an in silico trial, involving a simulated three-dimensional vessel of realistic shape. Then, we proceeded in vitro by acquiring volume data from an artificial capillary with multi-bifurcations made of poly(vinyl alcohol) by grayscale lithography whose diameter ranges from 0.5 to 2.0mm and with different flow velocities. Results show that our system successfully reconstructed the corresponding BVNs. ISCA

    Original languageEnglish
    Pages (from-to)203-210
    Number of pages8
    JournalInternational Journal of Computers and their Applications
    Volume21
    Issue number4
    Publication statusPublished - Dec 1 2014

    All Science Journal Classification (ASJC) codes

    • Computer Science(all)

    Fingerprint Dive into the research topics of 'Reconstruction and auto-correction of artificial capillary with flow directions from ultrasound volume data'. Together they form a unique fingerprint.

    Cite this