Skip to content

Advertisement

  • Poster presentation
  • Open Access

High resolution, free-breathing coronary artery imaging with >99% respiratory efficiency

  • 1,
  • 2 and
  • 1
Journal of Cardiovascular Magnetic Resonance201012 (Suppl 1) :P35

https://doi.org/10.1186/1532-429X-12-S1-P35

  • Published:

Keywords

  • Selective Excitation
  • Respiratory Efficiency
  • Corrected Dataset
  • Respiratory Motion Correction
  • Respiratory Position

Introduction

High-resolution coronary artery acquisitions are generally gated to end-expiration using navigators and have intrinsically low respiratory efficiency (RE), which is exacerbated by respiratory drift. Alternatively, epicardial fat can be used as a marker of coronary artery position in a 3D beat-to-beat non-model based subject-specific respiratory motion correction (B2B-RMC) technique[1]. We propose this technique can acquire high-resolution coronary artery images with approximately 100% RE.

Methods

In-plane right coronary images were acquired in 10 subjects on a Siemens 1.5 T Avanto scanner. For B2B-RMC, a 3D low resolution spiral dataset with fat selective excitation was acquired every cardiac cycle immediately before 2 interleaves of a 3D high-resolution spiral dataset with water selective excitation. A following navigator was used to reject data at extreme respiratory positions (>10 mm outside normal range). Beat-to-beat respiratory displacement of the coronary artery was determined from the low resolution images using localized 3D normalized sub-pixel cross-correlation of fat around the coronary origin (relative to end-expiration) and used to retrospectively correct the corresponding high-resolution interleaves. Navigator gated (5 mm window) 3D balanced steady-state free-precession (nav-bSSFP) with T2-prep and identical resolution to the B2B-RMC technique was chosen for comparison.

Analysis

For each B2B-RMC and nav-bSSFP dataset, a maximum intensity projection was generated with anatomy overlying the artery nulled. Average vessel diameter (full-width half maximum) and sharpness (inverse of 20-80% intensity distance[2]) were measured in the proximal (0-20 mm) and mid (20-40 mm) arteries.

Results

The RE was 99.7% (range 98.4-100%) and 40.7%, (range 33.2-53.6%) for B2B-RMC and nav-bSSFP respectively (paired t-test p < 0.0001). There was no significant difference in proximal or mid sharpness between the two methods (table 1), no significant difference between methods for mid diameters and a significant but insubstantial (0.15 mm) difference in proximal diameter which is possibly due to reduced vessel wall signal in the T2-prepared nav-bSSFP technique. Figure 1 shows B2B-RMC images optimally corrected for the proximal(a) and distal(b) vessel (RE = 99.3%) and the equivalent nav-bSSFP results(c) (RE = 41.0%).
Table 1

Average right coronary artery sharpness and diameter obtained using two imaging techiques

 

Proximal sharpness

(mm-1)

Mid sharpness

(mm-1)

Proximal diameter

(mm)

Mid diameter

(mm)

B2B-RMC (standard deviation)

1.00 (0.14)

1.01 (0.11)

2.85 (0.38)

2.85 (0.39)

nav-bSSFP (standard deviation)

1.08 (0.11)

1.05 (0.12)

2.70 (0.34)

2.80 (0.35)

Paired t-test

p-value

0.15

0.24

0.026

0.60

Figure 1
Figure 1

Curved planar reformat of 3D spiral images acquired with the B2B-RMC technique (0.7 × 0.7 × 3 mm resolution) in 302 cardiac cycles (93.3% efficient) over 20 mm of diaphragm motion, corrected for optimal proximal (a) and distal (b) vessel quality. For comparison the equivalent curved planar reformat of the 3D nav-bSSFP with identical resolution acquired in 612 cardiac cycles (41.0% efficient) is shown (c).

Conclusion

High-resolution coronary artery images were acquired with 99.7% RE using B2B-RMC in 10 subjects. Diameter and sharpness in the proximal and mid vessel were not substantially different to values obtained with nav-bSSFP acquired with mean RE 40.7%. For optimal visualisation, the B2B-RMC method requires cross-correlation of a local region of fat, as demonstrated in (a) and (b). Further work will merge images corrected for both proximal and distal vessel motion to generate a single corrected dataset.

Authors’ Affiliations

(1)
The National Heart and Lung Institute, Imperial College London, London, UK
(2)
Cardiovascular Magnetic Resonance Unit, The Royal Brompton Hospital, London, UK

References

  1. Keegan J: JMRI. 2007, 26: 624-10.1002/jmri.20941.View ArticlePubMedGoogle Scholar
  2. Li D: Radiology. 2001, 219: 270View ArticlePubMedGoogle Scholar

Copyright

Advertisement