Current limitations of coronary magnetic resonance angiography (MRA) include a suboptimal signal-to-noise ratio (SNR), which limits spatial resolution and the ability to visualize distal and branch vessel coronary segments. Improved SNR is expected at higher field strengths, which may provide improved spatial resolution. However, a number of potential adverse effects on image quality have been reported at higher field strengths. The limited availability of high-field systems equipped with cardiac-specific hardware and software has previously precluded successful in vivo human high-field coronary MRA data acquisition. In the present study we investigated the feasibility of human coronary MRA at 3.0T in vivo. Most recent studies suggest that free-breathing 3D coronary magnetic resonance angiography (MRA) at 1.5T is a valuable tool for the noninvasive assessment of significant proximal to mid coronary artery disease (1). Still, the ability to visualize and assess more distal or branching vessels is desirable, and quantitative grading of the stenoses remains an ultimate goal. Limitations of current coronary MRA approaches include residual motion and signal-tonoise (SNR) constraints. In theory, SNR is directly related to the strength (B 0 ) of the static magnetic field. Thus, improved SNR can be expected from the use of magnets with higher magnetic field strength, for which 3.0T systems have recently been approved for clinical use. However, together with the limited availability of higher-field systems equipped for cardiac applications, other potential impediments (2) include increased susceptibility artifacts, reduced T * 2 (3,4), increased T 1 , radiofrequency (RF) field distortions (5), and changed tissue dielectric constants or body dielectric resonances (6,7). Further, at higher field strengths, increased RF deposition may remove flexibility for general sequence design, and efficient myocardial motion suppression (reliable R-wave triggering) becomes more challenging due the amplified magneto-hydrodynamic (MHD) effects (8). Recent progress in 3.0T hardware (vector ECG (VECG) (8), short-bore magnets, dedicated surface receive coils (5), and body send coils) and the adaptation of cardiac-specific software (T 2 Prep (9) and 2D selective real-time navigators (10)) for 3.0T has made it possible to implement analogous 1.5T coronary MRA methodology on a whole-body 3.0T system. Therefore, we sought to investigate the feasibility of coronary MRA data acquisition at 3.0T. The first in vivo results obtained in nine consecutive healthy adult human subjects are presented. METHODSCoronary MRA was implemented on a commercial actively shielded whole-body compact magnet (Intera 3.0T, 157 cm magnet length, 60 cm internal diameter; Philips Medical Systems, Best, The Netherlands) equipped with a MASTER ® gradient system (30 mT; 150 mT/m/ms), a VECG module (8), and Release 8.1.3 cardiac scanner software. A prototype body send coil and a prototype sixelement cardiac synergy receive coil were used. Nine consecutive healthy adult human ...

No abstract

For many cardiac MR applications, respiratory bellows gating is attractive because it is widely available and not disruptive to or dependent on imaging. However, its use is uncommon in cardiac MR, because its accuracy has not been fully studied. Here, in 10 healthy subjects, the bellows and respiratory navigator (NAV) with the displacement of the diaphragm and heart were simultaneously monitored, during single-shot imaging. Furthermore, bellows-gated and NAV-gated coronary MRI were compared using a retrospective reconstruction at identical efficiency. There was a strong linear relationship for both the NAV and the abdominal bellows with the diaphragm (R 5 0.90 6 0.05 bellows, R 5 0.98 6 0.01 NAV, P < 0.001) and the heart (R 5 0.89 6 0.06 bellows, R 5 0.96 6 0.02 NAV, P 5 0.004); thoracic bellows correlated less strongly. The image quality of bellows-gated coronary MRI was similar to NAVgated and superior to no-gating (P < 0.01). In conclusion, bellows provides a respiratory monitor which is highly correlated with the NAV and suitable for respiratory compensation in selected cardiac MR applications. Magn Reson Med 65:1098-1103,

No abstract

Long scan times are still a main limitation in free-breathing navigator-gated 3D coronary MR angiography (MRA). Unlike other MRI applications, high-resolution coronary MRA has not been amenable to acceleration by parallel imaging techniques due to signal-to-noise ratio (SNR) concerns. In the present work, mitigating SNR limitations by the transition to higher static magnetic field strength is proposed, thus enabling scan time reduction by the parallel sensitivity encoding (SENSE) technique. The study reports the implementation and evaluation of free-breathing navigator-gated 3D coronary MRA with SENSE at 3T. Results from 11 healthy subjects indicate that the approach permits significant scan time reduction in MRA of the left and right coronary systems. Quantitative image analysis and visual grading suggest that two-fold scan acceleration can be accomplished at nearly preserved image quality. For an accurate localization of coronary artery disease, extensive portions of the left and the right coronary artery systems have to be imaged. Due to the tortuous orientation of the coronary arteries, multiple slices or 3D data have to be acquired to cover a coronary artery tree. Additionally, small vessel diameters necessitate submillimeter resolution and hence large data matrices. As a consequence, the mere amount of data required makes it difficult to complete data acquisition in a single breathhold (1,2). Hence, the scan may either be split into multiple breathholds (3,4) or performed with MR navigators (5,6) during free breathing.One of the techniques that have shown potential for the visualization of the coronary artery system is 3D coronary MR angiography (MRA) performed during free breathing and with navigator-gating and prospective real-time correction for respiratory motion (7). However, the short time window (80 -100 ms) of relative intrinsic cardiac rest during mid-diastole and the limited navigator efficiency (50% on average) makes scan times relatively long. The acquisition of a single high-resolution 3D stack, covering either the left or right coronary arterial system, takes ϳ10 min. This duration is quite long in terms of patient comfort, especially if we further consider the time required for localizer scans or a potential combination of the coronary MRA with other cardiac scans.The scan efficiency of navigator-gated acquisitions may be increased by combining navigator technology with sophisticated motion compensation strategies (8,9). As demonstrated by Jhooti et al. (8), k-space reordering permits increasing the navigator acceptance window and hence speeding up the data acquisition by a factor of 1.5 relative to a nonreordered method. Alternatively, or in addition, parallel imaging techniques such as sensitivity encoding (SENSE) (10) may be used, enabling scan time reduction by factors of 2 or even more. Like other parallel approaches, SENSE relies on signal acquisition with a receiver coil array. Reducing only the density of k-space sampling usually changes neither the imaging sequence itself nor the re...

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.

The usefulness of the proposed scan interleaving framework was demonstrated for image-based respiratory motion correction. It facilitated more direct comparisons of image navigator acquisitions with different k-space trajectories. Furthermore, we could demonstrate that spiral and Cartesian EPI navigators may be particularly suitable for image-based motion correction, as they provide improved motion correction and high navigator apparent signal-to-noise ratio while spending very little magnetization, thereby minimizing saturation effects.

Due to their relatively small size and central location within the thorax, improvement in signal-to-noise (SNR) is of paramount importance for in vivo coronary vessel wall imaging. Thus, with higher field strengths, coronary vessel wall imaging is likely to benefit from the expected "near linear" proportional gain in SNR. In this study, we demonstrate the feasibility of in vivo human high field (3 T) coronary vessel wall imaging using a free-breathing black blood fast gradient echo technique with respiratory navigator gating and real-time motion correction. With the broader availability of more SNR efficient fast spin echo and spiral techniques, further improvements can be expected.