Skip to content


Volume 18 Supplement 1

19th Annual SCMR Scientific Sessions

  • Oral presentation
  • Open Access

Free-breathing 3D whole-heart coronary mra using respiratory motion-resolved sparse reconstruction

  • 1, 2,
  • 3,
  • 1, 4,
  • 2, 4,
  • 2, 4,
  • 5,
  • 6,
  • 3,
  • 3 and
  • 2, 4
Journal of Cardiovascular Magnetic Resonance201618 (Suppl 1) :O105

  • Published:


  • Left Anterior Descend
  • Right Coronary Artery
  • Respiratory Motion
  • Left Main
  • Sparse Reconstruction


Navigator gating is commonly used to minimize respiratory motion in free-breathing whole-heart coronary MRA [1]. However, lengthy and unpredictable acquisition times remain a drawback. Respiratory self-navigation (SN) [2-3], conversely, enables 100% scan efficiency, but performs motion correction over a broad range of respiratory displacements, which can result in image artifacts. Here, we propose an alternative respiratory motion-resolved approach based on 3D radial phyllotaxis sampling, respiratory motion sorting and sparse reconstruction.


Examinations in N = 11 healthy volunteers (9 male, age: 29 ± 4 y) were performed on a 1.5T clinical MRI scanner (MAGNETOM Aera, Siemens Healthcare) with a prototype 3D radial phyllotaxis bSSFP sequence [4]: TR/TE 3.1/1.56 ms, FOV (220 mm)3, matrix 1923, voxel (1.15 mm)3, RF angle 115°, and receiver BW 898 Hz/Px. Using a respiratory signal directly extracted from the modulations of the k-space center amplitude within the radial imaging data [5], signal-readouts were grouped according to the respiratory state at which they were acquired (Fig. 1). The resulting series of undersampled respiratory states were reconstructed using an eXtra-Dimensional Golden-angle RAdial Sparse Parallel imaging (XD-GRASP) [6] algorithm, which exploits sparsity along the newly created respiratory dimension. Datasets for 4 respiratory states were reconstructed. Image quality of the end-expiratory phase was compared to 1D respiratory self-navigation in terms of vessel sharpness (VS) [7], visible length (VL) and diagnostic quality on a scale from 0 (non-visible) to 2 (diagnostic).
Figure 1
Figure 1

Example of results showing the improvements (arrows) between 1D self-navigation and XD-GRASP reconstruction.


Respiratory-resolved XD-GRASP reconstruction effectively suppresses respiratory motion artifacts (Fig. 1). Average VS and VL were always superior for the respiratory-resolved datasets, reaching statistical significance (p < 0.05) for the left main (LM), for the proximal and mid left anterior descending artery (LAD) (e.g. VS of mid LAD 40.8 ± 9.1% vs 34.9 ± 10.2%) and for the mid right coronary artery (RCA). Visualized length of LM+LAD was significantly increased as well. A total of 41/88 coronary segments were graded as diagnostic for 1D SN, while this ratio increased to 61/88 for the XD-GRASP reconstruction (Tab.1). The XD-GRASP reconstruction reached 100% diagnostic quality for LM, proximal-LAD, and proximal-RCA.


Instead of discarding data or enforcing motion models for motion correction, XD-GRASP makes constructive use of all respiratory phases to improve image quality, and achieves superior quality compared to 1D respiratory SN without the need for breath-holding, navigators, or complex 3D respiratory motion correction schemes. The phyllotaxis trajectory and XD-GRASP reconstruction provide a synergistic combination that may lead routine coronary MRA closer to clinical practice.
Table 1

Diagnostic quality grading of all coronary segments

Coronary Segment

1D Respiratory Self-Navigation

4-Phase X-D GRASP (End-exp)

Left Main

1.8 ± 0.4

2.0 ± 0.0*

LAD Prox.

1.6 ± 0.5

2.0 ± 0.0*


1.3 ± 0.6

1.4 ± 0.5

LAD Dist.

0.9 ± 0.5

1.3 ± 0.5

LCX Prox.

1.4 ± 0.7

1.4 ± 0.7

RCA Prox.

1.8 ± 0.4

2.0 ± 0.0


1.3 ± 0.5

1.7 ± 0.5

RCA Dist.

1.4 ± 0.7

1.7 ± 0.5

Total Diagnostic Segments

41/88 (47%)

61/88 (70%)

All values are expressed as mean ± one standard deviation

* Indicates statistical significance compared to 1D Respiratory Self-Navigation.

Diagnostic Grading: 0 = non-visible, 1 = visible but non diagnostic and 2 = visible and diagnostic

Authors’ Affiliations

Advanced Clinical Imaging Technology, Siemens Healthcare, Lausanne, Switzerland
Department of Radiology, University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland
Department of Radiology, Center for Advanced Imaging Innovation and Researc (CAI2R) and Bernard and Irene Schwartz Center for Biomedical Imaging, New York, NY, USA
Center for Biomedical Imaging (CIBM), Lausanne, Switzerland
Department of Radiology, Austin Health and The University of Melbourne, Melbourne, VIC, Australia
Division of Cardiology and Cardiac MR Center, University Hospital of Lausanne (CHUV), Lausanne, Switzerland


© Piccini et al. 2016

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.