Modelling of image-catheter motion for 3-D IVUS

Misael Rosales, Petia Radeva, Oriol Rodriguez-Leor, Debora Gil

Research output: Contribution to journalArticleResearchpeer-review

17 Citations (Scopus)

Abstract

Three-dimensional intravascular ultrasound (IVUS) allows to visualize and obtain volumetric measurements of coronary lesions through an exploration of the cross sections and longitudinal views of arteries. However, the visualization and subsequent morpho-geometric measurements in IVUS longitudinal cuts are subject to distortion caused by periodic image/vessel motion around the IVUS catheter. Usually, to overcome the image motion artifact ECG-gating and image-gated approaches are proposed, leading to slowing the pullback acquisition or disregarding part of IVUS data. In this paper, we argue that the image motion is due to 3-D vessel geometry as well as cardiac dynamics, and propose a dynamic model based on the tracking of an elliptical vessel approximation to recover the rigid transformation and align IVUS images without loosing any IVUS data. We report an extensive validation with synthetic simulated data and in vivo IVUS sequences of 30 patients achieving an average reduction of the image artifact of 97% in synthetic data and 79% in real-data. Our study shows that IVUS alignment improves longitudinal analysis of the IVUS data and is a necessary step towards accurate reconstruction and volumetric measurements of 3-D IVUS. © 2008 Elsevier B.V. All rights reserved.
Original languageEnglish
Pages (from-to)91-104
JournalMedical Image Analysis
Volume13
DOIs
Publication statusPublished - 1 Feb 2009

Keywords

  • Fourier
  • Intravascular ultrasound (IVUS)
  • Motion decomposition
  • Motion estimation

Fingerprint Dive into the research topics of 'Modelling of image-catheter motion for 3-D IVUS'. Together they form a unique fingerprint.

Cite this