Prospective respiratory motion correction for coronary MR angiography using a 2D image navigator

Markus Henningsson*, Jouke Smink, Reza Razavi, René M Botnar

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

43 Citations (Scopus)

Abstract

Respiratory motion remains the major impediment in a substantial amount of patients undergoing coronary magnetic resonance angiography. Motion correction in coronary magnetic resonance angiography is typically performed with a diaphragmatic 1D navigator (1Dnav) assuming a constant linear relationship between diaphragmatic and cardiac respiratory motion. In this work, a novel 2D navigator (2Dnav) is proposed, which prospectively corrects for translational motion in foot-head and left-right direction. First, 1Dnav- and 2Dnav-based motion correction are compared in 2D real time imaging experiments, by evaluating the residual respiratory motion in 10 healthy subjects as well as in a moving vessel phantom. Subsequently, 1Dnav and 2Dnav corrected high-resolution 3D coronary MR angiograms were acquired, and both objective and subjective image quality were assessed. For a gating window of 10 mm, 1Dnav and 2Dnav performed equally well; however, without any respiratory gating, the 1Dnav had a lower visual score for all coronary arteries compared with 10 mm gating, whereas the 2Dnav without gating performed similar to 1Dnav with 10 mm gating.
Original languageEnglish
Pages (from-to)486-494
Number of pages9
JournalMagnetic Resonance in Medicine
Volume69
Issue number2
Early online date23 Apr 2012
DOIs
Publication statusPublished - Feb 2013

Keywords

  • COMPENSATION
  • respiratory motion correction
  • MAGNETIC-RESONANCE ANGIOGRAPHY
  • 2D navigator
  • HEART
  • ALGORITHM
  • RECONSTRUCTION
  • ARTERY MOTION
  • prospective motion correction
  • coronary MR angiography

Fingerprint

Dive into the research topics of 'Prospective respiratory motion correction for coronary MR angiography using a 2D image navigator'. Together they form a unique fingerprint.

Cite this