Single breathhold noncontrast thoracic MRA using highly accelerated parallel imaging with a 32-element coil array

Abstract

Purpose: To evaluate the feasibility of performing single breathhold three-dimensional (3D) thoracic noncontrast MR angiography (NC-MRA) using highly accelerated parallel imaging.

Materials and methods: We developed a single breathhold NC MRA pulse sequence using balanced steady state free precession (SSFP) readout and highly accelerated parallel imaging. In 17 subjects, highly accelerated noncontrast MRA was compared against electrocardiogram-triggered contrast-enhanced MRA. Anonymized images were randomized for blinded review by two independent readers for image quality, artifact severity in eight defined vessel segments and aortic dimensions in six standard sites. NC-MRA and CE-MRA were compared in terms of these measures using paired sample t- and Wilcoxon tests.

Results: The overall image quality (3.21 ± 0.68 for NC-MRA versus 3.12 ± 0.71 for CE-MRA) and artifact (2.87 ± 1.01 for NC-MRA versus 2.92 ± 0.87 for CE-MRA) scores were not significantly different, but there were significant differences for the great vessel and coronary artery origins. NC-MRA demonstrated significantly lower aortic diameter measurements compared with CE-MRA; however, this difference was not considered clinically relevant (>3 mm difference) for less than 12% of segments, most commonly at the sinotubular junction. Mean total scan time was significantly lower for NC-MRA compared with CE-MRA (18.2 ± 6.0 s versus 28.1 ± 5.4 s, respectively; P < 0.05).

Conclusion: Single breathhold NC-MRA is feasible and can be a useful alternative for evaluation and follow-up of thoracic aortic diseases.

Figure 1

Schematic of NC-MRA using ECG-triggered, T2-prepared, Fat-saturated, segmented b-SSFP with the coil sensitivity and image data acquired at two different cardiac phases (early systole and mid diastole, respectively) in the same cardiac cycle within a single BH, in order to increase the acceleration efficiency and avoid the misregistration due to varying breath hold positions.

Figure 2

Diagram of the thoracic aorta demonstrates the segments used for the evaluation of image quality and the levels used for the measurement of aortic dimensions.

Figure 3

MR images in a 59-year-old patient suffering from aneurysm of the aorta root obtained with NC-MRA (Right) and CE-MRA (Left) sequence. The relative merits of each technique are: CE-MRA provides superior vascular delineation due to the enhancement by contrast; whereas the single BH ECG-triggered NC-MRA is relatively free from motion artifact, so cardiac morphology is more clearly visualized, with sharper delineation of the aortic root and better assessment of coronary artery origins, due to its much shorter acquisition window (110 ms vs. 216 ms; NC-MRA vs. CE-MRA, respectively).

Figure 4

A) MR images in a 79-year-old patient with tortuous thoracic aorta obtained with NC-MRA (Right) and CE-MRA (Left) sequence. The reconstruction planes clearly demonstrate sharper delineation of the aorta root and other aorta segments with high image quality for NC-MRA. B) Multiplanar reformatted images from NC-MRA (Right) and CE-MRA (Left) at the mid-descending aorta: sagittal, coronal and axial images.

Figure 5

Bland-Altman plot shows that overall measurements made with 3D NC-MRA and CE-MRA data sets were in good agreement with no significant difrerence in measured values. Bland-Altman analysis revealed a mean measured difference (NC-MRA minus CE-MRA) in aortic diameter between the two techniques of −0.073+/−0.144 cm upper and lower 95% limits of agreement = 0.210 and −0.357, respectively.