Purpose:To prospectively compare a new three-dimensional (3D) radial phase-contrast magnetic resonance (MR) angiographic method with contrast material-enhanced MR angiography for anatomic assessment of the renal arteries.
Materials and Methods:An institutional review board approved this prospective HIPAA-compliant study. Informed consent was obtained. Twenty-seven subjects (mean age, 52.6 years 6 20.5 [standard deviation]) were imaged with respiratory-gated phase-contrast vastly undersampled isotropic projection reconstruction (VIPR) prior to contrast-enhanced MR angiographic acquisition with a 3.0-T clinical system. The imaging duration for phase-contrast VIPR was 10 minutes and provided magnitude and complex difference ("angiographic") images with 3D volumetric (320 mm) coverage and isotropic high spatial resolution (1.25 mm 3 ). Quantitative analysis consisted of comparing vessel diameters between the two techniques. Qualitative assessment included evaluation of the phase-contrast VIPR and contrast-enhanced MR angiographic techniques for artifacts, noise, and image quality. Bland-Altman analysis was used for comparison of quantitative measurements, and the Wilcoxon signed rank test was used for comparison of qualitative scores.
Results:Phase-contrast VIPR images were successfully acquired in all subjects. The vessel diameters measured with phasecontrast VIPR were slightly greater than those measured with contrast-enhanced MR angiography (mean bias = 0.09 mm). Differences in mean artifact, quality scores for the proximal renal arteries, and overall image quality scores between phase-contrast VIPR and contrast-enhanced MR angiographic techniques were not statistically signifi cant ( P = .31 and .29, .27 and .39, and .43 and .69 for readers 1 and 2, respectively). The quality scores for the segmental renal arteries were higher for phase-contrast VIPR than for contrast-enhanced MR angiography ( P , .05).Although the noise scores were higher with phase-contrast VIPR than with contrast-enhanced MR angiography and were statistically signifi cant ( P , .05), the presence of noise did not interfere with the ability to interpret the images.
Conclusion:Isotropic, high-spatial-resolution, unenhanced MR angiography of the renal arteries is feasible with 3D radial undersampling.q RSNA, 2010 1