Coronary magnetic resonance angiography at 7 Tesla: a quantitative comparison with results at 3 Tesla

Introduction

Commercial 7 Tesla(T) systems have recently become available for human use. Our initial results have shown feasibility of right coronary artery(RCA) magnetic resonance angiography(MRA) at 7 T[1]. Since then, we have optimized the RF coil design, resulting in significantly improved image quality at 7 T.

Purpose

To quantitatively compare same-subject signal-to-noise(S/N), coronary vessel length, vessel sharpness, acquisition time and navigator efficiency of right coronary MRA at 7 T and 3 T.

Methods

Eight healthy adult subjects (mean age 23 ± 3 years, 5 men) underwent vector ECG-triggered, navigator gated and corrected free-breathing 3D MRA of the RCA at 7 T and 3 T. At 7 T a quadrature loop-pair (two 15 cm elements) was constructed, and the 2D selective respiratory navigator was localized at the lung-heart interface. At 3 T, a commercial 6-element cardiac receive array was used with body-coil transmit. A 3D segmented k-space gradient echo technique was combined with spectrally selective adiabatic inversion-recovery magnetization fat saturation. At 3 T, coronary MRA were obtained with the navigator localized at the lung-heart interface and at the lung-liver interface, respectively. The scan parameters at both field strengths were as similar as possible (Table 1). The S/N in the bloodpool of the aortic root near the RCA offspring, coronary vessel length, vessel sharpness, acquisition time and navigator efficiency were compared using Wilcoxon matched-pairs test. For visualization, reformatting with the 'Soapbubble'-tool was utilized.

Table 1 Scan parameters

Results

In Figure 1, RCAs obtained in the same subject at 7 T(Figure 1a) and 3 T (Figure 1b) are shown. The corresponding quantitative findings averaged over all subjects are listed in Table 2. The S/N was very similar at both field strengths and visible RCA vessel length was slightly lower at 7 T, but not statistically significant. Vessel sharpness obtained at 7 T was significantly higher compared to that at 3 T (p < 0.03). The scanning time for all 3D acquisitions was identical. There was a tendency for reduced navigator efficiency at 7 T which was not statistically significant, however. These findings apply for both the diaphragmatic navigator as well as for that localized at the heart with a trend for a lower S/N at 3 T for the navigator localized at the lung-heart interface.

Table 2 Quantitative findings, *significantly different from 7 T (p < 0.05)

Figure 1

Conclusion

Using a quadrature transmit/receive coil at 7 T, it was shown that 7 T coronary MRA image quality has already begun to approach that at 3 T while vessel sharpness is already significantly improved. With future incorporation of technology such as larger transmit arrays we anticipate that the image quality at 7 T will continue to improve at rapid pace.