Breath-hold 3D MR angiography of the renal vasculature using a contrast-enhanced multiecho gradient-echo technique

Significant evolution of contrast-enhanced MR angiography for evaluating vascular diseases in the abdomen has occurred during the past several years. The state-of-the-art gradient-echo imaging technique employs a short echo time (TE) and a short repetition time (TR) for rapid vascular imaging with contrast-enhanced MR angiography. However, because of the short TR (< or = 3-8 msec), the background stationary tissue becomes saturated, with resultant poor contrast resolution of visceral organs. The authors present a new approach to vascular imaging using a multiecho gradient-echo technique with a TR sufficiently long (41 msec) to image the renal vasculature and parenchyma without background tissue suppression.

Twenty-four partitions (3D slab thickness = 72 mm) with an in-plane resolution of 224 x 256 were obtained in 21 seconds. Three measurements were performed with an interscan delay of 8 seconds. In the pulse sequence, the partition loop is defined as the innermost loop, in which Kz views are acquired centrically for a fixed Ky, followed by Ky views in a conventional linear or sequential order. The partition encodings are segmented to permit multiple encodings in which two TR loops were used to span a total of 24 echoes with 12 along the positive and 12 along the negative direction in k space. A large bandwidth of 650 Hz/pixel was used to keep the echo train length short, with an echo spacing of 1.86 msec. A frequency-selective fat saturation pulse was placed before slab-selective excitation. The other parameters in the pulse sequence were TR/TE/flip = 41/2.2/45; the field of view was 360 to 390 mm. Maximum intensity projections of each 3D contrast-enhanced measurement were performed. The vascular-to-background contrast, bowel-related magnetic susceptibility artifact, and background stationary signals were subjectively graded. The authors examined the utility of this technique in 16 randomly selected patients (3 normal, 13 abnormal) with varied renal vasculature and parenchymal abnormalities. Results were confirmed with conventional x-ray angiography, surgery, or clinical follow-up.

Vascular-to-background contrast was graded as very good (grade III/III) in all cases. The bowel-related magnetic susceptibility artifacts were not considered significant. Background visceral organ soft tissue contrast was not suppressed and was graded as good (grade III/III) in all cases. Eight hemodynamically significant (> 50% diameter) stenoses in seven patients were accurately assessed (one with fibromuscular dysplasia). Three patients with renal masses (two with renal cell carcinoma and one with renal lymphoma) were accurately assessed for arterial anatomy and venous extension. Other renal venous abnormalities seen were retroaortic renal vein (n = 1), chronic occlusion (n = 1), and accessories (total of five) (n = 1).

Rapid breath-hold contrast-enhanced MR angiography of the renal vasculature with a multiecho gradient-echo using a long TR depicted the renal vasculature with high vessel-to-background contrast without significant bowel-related susceptibility artifact and without background visceral organ tissue signal suppression, resulting in high background soft tissue contrast resolution.