Few doubt the ability of 3T to reveal exquisite detail on brain MR imaging. The argument for higher field strength scanning of small joints also appears cut and dried. The true test of 3T, however, is whether it can excel in areas of interest to general radiology. This level of acceptance may only be a matter of time, according to evidence gathered from the maturing technology.
Three-T body MRI has been relatively slow to take off. Get the technology and protocols right, however, and the improvements could be more dramatic than those witnessed in neuroradiology, said Dr. David Bluemke, director of MRI at Johns Hopkins University.
"The area where, in general, we are missing signal is body MRI. This has most to gain from the higher signal-to-noise ratio, but it is also one of the most difficult areas to implement 3T techniques," he said.
Interest in 3T body MRI is being fueled by the optimization of sequences and the greater availability of coils and pads to increase signal and counter artifacts. These advances are enabling images to overcome hurdles that had previously loomed large.
Papers demonstating the value of 3T for MR cholangiopancreatography (MRCP) are beginning to appear, said Dr. Elmar Merkle, director of body MRI at Duke University Medical Center. Limitations in spatial resolution and SNR at 1.5T have restricted the value of MRCP, but moving to 3T has traditionally produced artifacts. New sequences becoming available at 3T help to resolve these problems (see figure, page 64). This capability could allow the noninvasive imaging method to be used to evaluate intra- as well as extrahepatic biliary anatomy.
Current areas of activity for abdominal radiology researchers include blood oxygen level-dependent, perfusion, and dynamic contrast-enhanced imaging, said Dr. Jeffrey Weinreb, director of medical imaging at Yale University School of Medicine. Such techniques are expected to aid oncology, for instance, by depicting angiogenesis.
Weinreb also predicts considerable growth in 3T prostate imaging and spectroscopy. Yale radiologists are currently working without an endorectal coil for their 3T scanner, so they have yet to test the value of higher SNR for prostate MR studies.
But the need for such a coil for prostate MR is debatable, given the higher SNR. Arguments for and against the continued use of these imaging aids have yet to be resolved. The coils are uncomfortable for patients and can be tricky to position. They are unlikely to be abandoned, however, without conclusive proof that coil-less 3T techniques could provide comparable diagnostic information.
The latest research results indicate that 3T MRI performed with an endorectal coil offers the best option for staging prostate cancer, said Prof. Jelle Barentsz, chair of radiology research at the University Medical Center St. Raboud in Nijmegen, the Netherlands. Localization of pathology, where high resolution is less critical, may be possible without the coil.
"This is provided you use dynamic contrast-enhanced MRI and spectroscopy as well," he said.
Higher field strength magnets offer advantages for more precise staging and localization with a whole range of abdominal cancers, according to Barentsz. One set of 3T prostate images acquired at UMC St. Raboud, for example, revealed the patient's rectal cancer with superb resolution. Individual cases of cervical cancer, renal cell carcinoma, and liver metastases have also been well demonstrated at 3T at the Nijmegen center.
The added signal may be of benefit in the emerging area of lymph node imaging with ultrasmall superparamagnetic iron oxide contrast. Agents such as Sinerem or Combidex are absorbed less readily by metastatic nodes than by normal lymph nodes, allowing the two to be distinguished through signal differences on imaging. Switching to 3T means that smaller (3 mm to 4 mm) nodes can be classified with the same accuracy as larger nodes (5 mm and over) at 1.5T, Barentsz said. Data being gathered at Nijmegen indicate good results in both the prostate and pelvis.
EMERGING AREAS
Breast imaging is another area where 3T is only beginning to capitalize on its potential. The relative scarcity of centers using 3T scanners for breast MR is due to a lack of dedicated coils and the historic association of breast imaging with mammography, according to Prof. Christiane Kuhl, a professor of radiology at the University of Bonn. Yet contrast-enhanced breast MRI would show benefits from 3T's higher SNR.
Breast MRI requires accurate depiction of very fine anatomic details and capture of short-lived tumor enhancement. The initial contrast between cancers and surrounding parenchyma quickly fades and can be canceled out in two or three minutes at 1.5T.
"We need to image fast and with a very high spatial resolution," Kuhl said.
The accuracy of 3T breast MR studies is further boosted by the addition of MRS, she said. Research indicates that sizable choline peaks make a good independent marker of malignancy. Spectroscopy data collected at 1.5T are of little diagnostic value, however, because of the large voxel size. Sampling areas set to assess small lesions also include large areas of normal tissue, which has a significant influence on the MRS readout. Switching to 3T means voxel size can be reduced, and radiologists can be more confident of negative findings.
Spectroscopy forms a significant part of the 3T breast MR research at New York University, said Dr. Linda Moy, an assistant professor at NYU Medical Center. Radiologists use research time to investigate varying contrast doses and refine their standard T1- and T2-weighted imaging sequences. Future plans include the addition of perfusion and diffusion imaging to breast MR protocols.
Many 3T users have shied away from cardiac imaging applications. Motion artifacts caused by respiration and the heart's own beating are exacerbated with the higher SNR.
"The heart moves in a craniocaudal direction by several centimeters. If you want to do cardiac imaging, you need to have some sort of advanced artifact suppression technique," Kuhl said.
Researchers at Bonn are already gaining some success with their strategies for 3T coronary MR angiography, but the real gain for cardiac MR will have to come from 3T perfusion imaging, she said.
Meanwhile, radiology researchers at the University of California, Los Angeles are aiming to overcome difficulties with cardiac cine imaging at higher field strengths. Cine imaging protocols followed at 1.5T tend to use steady-state free precession. Transposition of the same sequences to 3T can cause problems with power deposition and off-resonance effects, said Dr. Paul Finn, a professor of radiology and medicine at the UCLA Cardiovascular Research Laboratory. Keeping the specific absorption rate within sensible limits means either increasing the repetition time and/or lowering the flip angle. Three-T cine imaging can be plagued by dark banded off-resonance artifacts caused by magnetic inhomogeneities.
Finn's team is investigating an alternative approach to 3T cardiac cine imaging using intravascular contrast agents now under development. Initial results in pigs have been extremely encouraging, he said.
