3T disappoints in MRA diagnosis of coronary disease

May 11, 2005
NeedsFixing
NeedsFixing

A head-to-head clinical comparison of 3T and 1.5T MRI indicates that the higher field strength produces no better than equivalent results for the diagnosis of coronary artery disease.

A head-to-head clinical comparison of 3T and 1.5T MRI indicates that the higher field strength produces no better than equivalent results for the diagnosis of coronary artery disease.

A study of 18 patients conducted by Dr. Torsten Sommer, director of cardiovascular imaging at the University of Bonn in Germany, found that the standard Navigator-gated 3D gradient-echo sequences performed at 3T yielded specificity and sensitivity rates of 82% and 89%, respectively, for detection of coronary stenoses. Sequences optimized for 1.5T imaging generated specificity and sensitivity rates of 82% and 88%.

Rather than an expected doubling of signal-to-noise ratio, the 3T platform actually generated 30% more SNR, according to Dr. Matthias Hackenbroch, a radiology resident who presented the paper at the SCMR meeting in January. Doubling the magnetic induction rate raised the contrast-to-noise ratio by 22%.

Dr. Matthias Stuber, who also spoke at the SCMR meeting, blamed the disappointing results on 3T's problems with specific absorption rate and B0 and B1 inhomogeneity problems. Suboptimal SAR and B0 homogeneity compromise the quality of steady-state free precession imaging. Some studies have shown a 100% difference in B1 homogeneity across the field-of-view. This degrades the performance of pulse sequences that use high flip angles, such a T2 Prep, an important sequence for contrast-enhanced coronary MR angiography.

Help is on the way, however, from innovative strategies discussed at the meeting. New 3T shimming algorithms compensate for B0 inhomogeneity and result in notable image quality improvements, said physicist Michael Schar from the University of Zurich in Switzerland, who developed the algorithms.

New adiabatic T2 Prep pulses have been equally adept at reducing artifacts associated with B1 inhomogeneity. Stuber, an associate professor of radiology at Johns Hopkins University, displayed images from his lab that characterized right coronary artery segments as long as 12 cm and arterial branching on the right and left coronary arteries that could not be observed with conventional T2 Prep sequences.

Resolution will improve even further with the commercialization of massively parallel imaging techniques such as those developed by Dr. Daniel K. Sodickson, director of biomedical imaging research at Beth Israel Deaconess Hospital in Boston. High-resolution coronary artery flow images can now be acquired at 0.7 x 0.7 mm with a temporal resolution as fast as 15 msec. Under experimental conditions, 17 pixels have been acquired across a cross-section of an RCA, more than a four-fold improvement over conventional techniques.

Another two years may be needed to port all cardiac sequences from the 1.5T to the 3T platforms, but Dr. Gerald Pohost, chief of cardiovascular medicine at the University of Southern California, predicts that the theoretical advantages of 3T will be realized.

"3T is the next frontier," he said.