Evaluation and comparison of ECG-gated techniques at 1.5 T for contrast enhanced MR angiography of the thoracic aorta

Background

3D T1-weighted contrast-enhanced MRA (CE-MRA) is routinely used for non-invasive evaluation of the thoracic aorta. However, competing demands of high spatial resolution and fast (breath-hold) acquisition often preclude ECG-gating, leading to motion artifact at the aortic root. Standard Cartesian-sampled ECG-gated CE-MRA acquires 1 partition per heartbeat resulting in long scan times, often exceeding breath-hold capabilities. We evaluated image quality and diagnostic capabilities of a novel ECG-gated CE-MRA utilizing alternative Cartesian k-space sampling whereby adjacent k_y and k_z points are acquired in a zigzag pattern (Z-MRA) to improve scanning efficiency and co-ordinate contrast timing with optimal cardiac phase acquisition.

Methods

42 patients (12 females, mean 52y) were enrolled and underwent CE-MRA at 1.5T (Avanto, Siemens Healthcare) using a two-injection protocol with standard ECG-triggered CE-MRA (S-MRA) and zigzag (Z-MRA) ECG-gated CE-MRA (IPR #573: Siemens Healthcare, Germany) performed in a randomized order following 0.15mmol/kg gadolinium contrast. S-MRA parameters were: TR 2.7/TE 0.9, FA 17°, FOV 400mm. Z-MRA parameters had matched spatial resolution and FOV with other parameters: TR 2.6ms/TE 0.9ms, FA 20°, time to center (TTC) approximately 4.5s, TTC per heartbeat (k_y=0) acquired on average 566ms post-trigger, 2-3 k_z loops per heartbeat (heart-rate dependent). A parallel imaging factor of 2 was used for both. Two physicians independently reviewed the images. Ten arterial segments were graded for image quality (IQ), artifacts, vascular contrast, pathology and diagnostic confidence.

Results

1680 segments (840 x 2 readers) were evaluated. No scans were considered non-diagnostic. Average scan time was significantly longer with S-MRA compared to Z-MRA (52.4 vs. 17.9 sec, p<0.001). Overall image quality was similar for S-MRA compared to Z-MRA (Table 1). Sinus and sinotubular junction IQ and artifact scores were significantly superior for S-MRA, but beyond the ascending aorta, IQ and artifacts scores were significantly superior for Z-MRA. Vascular contrast was significantly superior at all segments for Z-MRA. Overall diagnostic confidence was significantly better for S-MRA, mainly due to difficulty discerning pathology at the aortic root.

Table 1 Overall comparison scores for both sequences.

Conclusions

ECG-gated Z-MRA is feasible for diagnostic evaluation of the thoracic aorta with significantly superior vascular contrast and comparably lower breath-hold times. However, motion artifact at the aortic root led to decreased diagnostic confidence at these segments for Z-MRA. Z-MRA provides a reasonable alternative to S-MRA, particularly for patients with limited breath-hold capabilities or in whom we want to limit contrast dose. Further optimization of Z-MRA k-space sampling strategies are needed to improve overall diagnostic performance.

Funding

N/A.