February 16, 2006
High-Field MRI — The Pull of Big Magnets
Magnetic resonance machines are continuously being pushed toward higher magnetic fields in hopes of developing ever-better images without ionizing radiation.
Most clinical magnets today operate at 1.5 Tesla (T)—up from 1T or less when they were introduced for medical imaging in the 1980s. While 1.5T is the predominant field strength sold and installed today, the 3T magnet continues to gain market share, says Nancy Gillen, BA, RT(R)(MR), vice president of the MR division of Siemens Medical Solutions.
In 1998, the first clinical ultra high-field MRI scanner—operating at a field of 8T—at Ohio State University (OSU) in Columbus produced images of the human head. Following this successful launch into the medical imaging scene, the OSU group published numerous safety studies, says Alayar Kangarlu, PhD, a member of the 8T team at that time.
The studies, which reported no adverse health effects on animal and human subjects, helped the FDA in July 2003 to raise the limit of field strength associated with no significant risk to 8T. “That created the chance for more such devices being installed by other institutions,” Kangarlu says.
The earliest ultra high-field scanners were built by researchers with components and magnets from various providers, says John Patrick, PhD, MBA, director of business development for Philips Medical Systems. But manufacturers have since developed integrated ultra high-field systems with capabilities previously only available on clinical systems.
Today, a small number of facilities in the United States and abroad have acquired magnets of ultra high-field strengths. For example, last year, New York University’s (NYU) School of Medicine powered up its 7T magnet built by Siemens, and Philips installed a 7T system at its corporate facility in Cleveland.
Also, the ultra high-field magnets in use today are for research, mostly of the brain.
“Right now, there are no 7T scanners that are clinically marketed,” says Greg Sorensen, MD, codirector of the Athinoula A. Martinos Center for Biomedical Imaging, in Charlestown, Mass., which has had a 7T Siemens magnet since 2002. “No manufacturer has yet requested FDA approval for a 7T device—although they could.”
Throughout the history of MRI, many have taken a cautionary approach about the viability of higher field strength, Kangarlu says. “In fact, early in the history of MRI, even some experts expressed doubt whether MRI scanners could surpass the 1 Tesla landmark.” Kangarlu says the early skepticism was natural because of two distinct concerns: safety and radio frequency (RF) engineering. “In every state of MRI, the trend has been such that, as the safety issue is addressed, a flurry of work is launched in RF coil design, pulse sequences, and gradient engineering, contributing to the commercial feasibility and medical viability of that field strength. Ultra high field is no exception from this general rule.”
In the past few years, engineers have used a technique called parallel imaging to address a major technical difficulty of the higher-field systems. Parallel imaging uses arrays of multiple receiver coils to reduce acquisition time.
Cost is yet another issue with ultra high-field magnets. A rule of thumb used to be facilities could expect to spend $1 million per Tesla—thus a 1.5T scanner would be approximately $1.5 million, a 3T scanner $3 million, etc. “That’s more or less true plus or minus the siting,” Sorensen says.
However, Patrick says that ratio holds until approximately 7T. “From 7T to 9.4T, you’re looking at an increase of about $8 million. It’s an exponential increase in cost with diminishing returns.”
Above $10 million, the ratio is more like $10 million per Tesla, and so a 15T magnet would probably cost roughly $150 million, “and no one has built one that big for humans,” Sorensen says. Smaller research lab magnets have generated magnetic fields in the neighborhood of 14T.
Given the technical challenges and the cost, why is there a push for even stronger magnets? The answer, say those involved, is to be able to see the workings of the human body at levels that have never been seen before. Being able to do so could mean more personalized healthcare and earlier intervention to prevent disease.
With the current 1.5T scanners, researchers are limited to imaging hydrogen, the most abundant nucleus in the human body. (Water, composed of hydrogen and oxygen, makes up roughly 70% of the body.) As a result, only anatomical changes can be detected and monitored. Higher-field magnets allow the detection of many more key compounds in the body such as carbon, phosphorus, sodium, oxygen, and nitrogen.
“You start to get into areas where you can do meaningful imaging with meaningful resolution on those alternate nuclei,” adds Patrick.
With more powerful magnets, researchers can find the other signals buried beneath the water signal. They are more distinct because they no longer overlap, explains Robert Grossman, MD, chairman of the department of radiology at the NYU School of Medicine.
To date, most research being able to detect signals from other nuclei has been concentrated on the brain and neurological diseases, including multiple sclerosis, epilepsy, Parkinson’s, Alzheimer’s, and stroke.
Grossman’s NYU laboratory has identified the signature of a chemical called N-acetyl aspartate in the brains of people with multiple sclerosis. Early studies suggest that patients with an imbalance in this chemical may benefit from aggressive therapy.
A group of researchers from Bonn, Germany, were in Cleveland recently using Philip’s 7T system. “They set up protocols and obtained the images of multiple sclerosis at 7T,” Patrick says.
When a patient has an acute stroke, “the question you need to answer is whether the tissue behind the blood clot in the brain is already dead,” he says. “If it is, you don’t want to thrombolyze the clot, because it’s going to bleed and produce a bigger problem. But if it’s alive you should try real hard because you can save it. Sodium imaging can answer the question whether the tissue is alive or dead and, at ultra high-field, you can do it with greater sensitivity and decreased acquisition time.” And, in many cases, time is a factor.
Sodium imaging done at ultra high-field may also provide answers to the effectiveness of drug treatments in brain cancer patients, Thulborn says.
Drugs and treatments are designed to work in the majority of patients, but everyone is different. Sodium imaging could provide information that would allow physicians to tailor therapies to each person’s own body chemistry and brain function, Thulborn says.
Currently, physicians must wait six months to image the tumor again and determine anatomically whether therapy has made it shrink or had no effect. However, after several doses of therapy, sodium imaging could be done to determine whether the tissue is dying, similar to how the PET researchers studying that modality to assess treatment effectiveness. “The goal here is to answer the question, ‘Did you kill the cells?’ Sodium imaging actually allows you to answer that and it allows you to do so in a timeframe that matches whether you should change the therapy after three doses rather than after 20,” Thulborn says.
Still another area where ultra high-field MRI may prove useful is in planning for brain surgery. Scientists know generally where speech, sensation, memory, and other functions occur in the brain. However, the exact locations of these activities vary from person to person. Employing functional magnetic resonance imaging (fMRI) protocols, researchers can actually see metabolic changes that take place in the active parts of the brain. They can use stimuli such as noise, light, or a finger prick to cause and track the changes in blood flow while the patient is in the magnet.
Some imaging capability super-strength magnets promise is already available with PET and other imaging modalities. Researchers are investigating the MRI alternatives to avoid radiation exposure. Studies could be repeated with MRI as often as necessary without worrying about the patient reaching his or her accumulated lifetime dose of radiation, Sorensen says.
However, researchers don’t really see high-field MRI replacing existing imaging procedures. Rather, they see it complementing them.
“I think the 7T will be a niche for some time to come,” Sorensen says. “There are some conditions we can see at 1.5T just fine for today’s treatments. We don’t yet need more precision for some diseases. Medically, the extra information is not necessarily needed. I don’t think 3T or 7T scanners are going to rapidly take over the world for no other reason than they are simply more expensive to build and maintain, but I do suspect there will be growth.”
Some researchers expect that what they learn from 7T and higher scanners may ultimately translate to lower field magnets and thus be more commonly available. Because the resolution is much higher at 7T, researchers can see things they can’t at 1.5T or 3T, Patrick says. But once they know they are there and where to look, they may be able to find them as well on their 3T scanners.
“It may be that we take the information that is available at very high sensitivity and filter it down to the 3T environment through the use of enrichment and hyperpolarization rather than saying everyone has to work at ultra high-field,” Thulborn agrees.
Soon ... Not Really
“With a 7T, if you want to fairly closely shield the magnetic field, you need 440 tons of steel around it,” Patrick says. “For a magnet of equivalent size operating at 9.4T, you’re looking at close to 800 tons of steel where with 3T scanners, the magnets are actively shielded so you don’t need any steel at all.”
UIC’s 9.4T magnet, which took three years to build, required 520 tons, Thulborn notes. Twenty-five flatbed trucks delivered the 420 tons of steel NYU needed to surround its superconducting magnet.
Also, despite the studies, there is still some unease about safety. Studies of body temperature, electrocardiograms, heart rate, respiratory rate, blood pressure, and blood oxygenation while patients are in high-field magnets have yet to discover any adverse effect on human volunteers, Kangarlu says. However, he says, “further work to detect such effects is warranted.”
Thulborn, whose head has been scanned at 9.4T, says many people ask him about the biohazards of 9.4T. Heating is not the issue many people suspect it may be. “We have pulse sequences that are very energy-efficient so there is no sense of heating at all,” he says.
Thulborn has found the solution to these minor issues is to move into the magnet at a slow, fixed rate. “We found if you move someone in, and they get a metallic taste and you stop moving them, the taste disappears in a few seconds and you can continue to move them. You just have to move slowly.”
Kangarlu, who currently is the head of MRI Physics at Columbia University Department of Psychiatry and New York State Psychiatric Institute, says the higher the strength of the magnet, the stronger the interaction with the human body and greater the need for understanding its safety challenges.
Staff and patients have always had to be careful not to bring ferromagnetic objects into MRI suites. The force of the magnet can turn such objects—paper clips, hair pins, etc—into dangerous projectiles. “Regarding the gravity of this aspect of MRI scanners and their life-threatening potential it warrants a special attention and systematic training of the MRI staff to heighten their awareness of the projectiles’ effect in high-field MRI scanners,” Kangarlu says.
Over the next decade, Kangarlu expects high-field magnets will offer numerous exciting applications, including microscopic fiber tracking, neuronal fMRI, in vivo metabolite concentration measurements, and molecular imaging.
“It is up to the MRI scientists and engineers to make a special endeavor to acquire more insight in the health risks of high-field scanners while these new superior applications are being developed,” says Kangarlu.
Beth W. Orenstein of Northampton, Pa., is a freelance health writer
and a regular contributor to Radiology Today.
3801 Schuylkill Rd • Spring City, PA 19475
Publishers of Radiology Today
All rights reserved.