Novel techniques combined with MRA address cardiac, respiratory motion

Jan 16, 2001

In order to circumvent the problems associated with respiratory motion in cardiac imaging, two international teams of researchers are combining MR angiography with unique imaging algorithms that require either a single breath-hold or no breath-hold.

Both groups presented their work at the RSNA conference in November. First was Dr. Thomas Hany from the Institute of Diagnostic Radiology at the University Hospital Zurich in Switzerland. His co-authors on the study were from the University of Wisconsin in Madison.

Hany discussed time-resolved visualization of the coronary arteries with contrast-enhanced, 3-D undersampled projection reconstruction (PR) MRA. The technique requires a single breath-hold using an intravascular contrast agent.

"The problem that we have with imaging the coronary arteries is that we have respiratory movement as well as cardiac motion. Using the breath-hold technique, we can compensate for that and we can try to time-resolve the images," Hany said.

In this animal model, the coronary arteries of 10 pigs were scanned on a 1.5-tesla MR unit using a 3-D PR sequence. The coronary arteries were localized using a 2-D cine sequence. For the contrast-enhanced scan, an ECG-triggered undersampled PR technique with multiphase acquisition was used with the following parameters: 64 projections, 256 readout matrix, a field of view of 28 cm, and 14-slice encode values zero-filled to 28 slices with a slice-thickness of 2 mm.

At a heart rate of 120 beats per minute, the scan duration was 28 seconds with the single breath-hold lasting from 30 to 35 seconds, Hany said. The technique allowed for the acquisition of six to eight identical 3-D volumes throughout the cardiac cycle.

The left coronary (LCA), left circumflex (LCX), and right coronary (RCA) arteries were included in three separate, targeted 3-D volumes performed in the steady state at six to 10 minutes after intravenous injection of 0.1 mmol per kilogram body weight of Gadomer 17, (Schering, Berlin, Germany), a dendrimeric, gadolinium-based intravascular contrast medium.

X-ray angiography was also performed for comparison purposes.

According to the results, the duration of contrast enhancement after a single injection was sufficient for visualization of all three arteries. The average measured length of the LCA was 62 mm, 41 mm for RCA, and 43 mm for LCX. The average diameter of the mainstems was 2.3 mm with good correlation on x-ray angiography. No significant enhancement of the myocardium was detected.

"The undersampled PR allows faster data acquisition but results in lower signal-to-noise ratio, which can be resolved by using an intravascular contrast agent," Hany said. Visualization in a cine display means all coronary arteries are visible throughout the cardiac cycle.

In a second presentation, Dr. Christian Weber from the University Hospital in Hamburg, Germany discussed the results of a study comparing selective coronary angiography to 3D MR angiography with a motion-adapted gating technique.

"Motion-adapted gating technique is a navigator-based, free-breathing, high-resolution 3-D technique using the math algorithm for respiratory motion compensation," Weber said. "The math algorithm is based on k-space weight that enables an instant and automatic analysis of respiratory motion in real time, and it changes the gating term without the necessity of user interaction."

For this study, 11 patients and four healthy volunteers with an average age of 61 were scanned with a 1.5-tesla Gyroscan ACS-NT (Philips Medical Systems, Shelton, CT) equipped with a PowerTrak 600 gradient system. ECG-triggered, respiratory motion-gated, 3-D, turbo field-echo (TFE) sequences were acquired. The three main coronary arteries and left main (LM) coronary arteries were evaluated, and three blinded investigators performed a qualitative analysis. Visibility was graded on a zero-to-four scale, with one representing insufficient visualization, and four considered excellent visualization.

In the patient population, 62 out of 88 coronary arteries were defined as adequately visualized, with a rating between two and four. In the volunteer group, 22 out of 32 coronary arteries received the same score. Visibility was considered excellent for proximal RCA, good for LM, proximal left anterior descending, proximal LM, and middle RCA.

In terms of stenosis, "selective coronary angiography could visualize 88 coronary artery segments with 16 stenoses of more than 50%," Weber said. "[With the study technique], we adequately visualized 62 of 88 coronary arteries and we found 14 true-positive stenoses of more than 50%."

While Weber concluded that their technique showed great promise for noninvasive imaging of the coronary arteries with adequate image time and patient comfort, session moderator Dr. Charles Higgins cautioned against one pitfall.

"You were looking at detecting greater than 50% stenosis, which is obviously important," Higgins said. "With MRA, the pitfall is differentiating between occlusions and high-grade stenoses. Turbulent flow across the high-grade stenosis could mimic occlusions because of the signal dropout."

By Shalmali Pal
AuntMinnie.com staff writer
January 17, 2001

Click here to post your comments about this story. Please include the headline of the article in your message.