SIEMENS

Imagine if your car had no visible controls, operated on an obscure mixture of biochemical substances, and yet somehow managed to take you exactly where you wanted to go. Our minds are a lot like such a vehicle. At any given instant, innumerable processes are at work in our heads that we are neither aware of nor have any objective way of controlling. That's fine most of the time. But when trouble crops up—severe depression, anxiety, obsessive-compulsive disorders, or one of a vast range of other potentially debilitating psychological conditions—the only controls we have access to are counseling sessions and medications with mile-long lists of side effects.

Now, a new road to managing such conditions has opened up—one that empowers patients while potentially obviating many expensive medical therapies. Thanks to a close research collaboration between Siemens and the Massachusetts Institute of Technology (MIT), scientists at MIT's McGovern Institute for Brain Research in Boston are in the process of developing tools that hold the promise of allowing many patients to manage their conditions by learning to voluntarily modulate otherwise inaccessible processes.

Can You Modulate Your Amygdalae? Take depression, for instance. Thanks to a substantial body of knowledge, MRI can now support diagnosis of this condition. “We know that if patients with depression respond to treatment, their amygdalae—two almond-shaped structures located deep beneath the temples—will demonstrate a reduced level of activity,” says Dr. John Gabrieli, Grover Herman Professor of Health Sciences, Technology and Cognitive Neuroscience at MIT and Director of the Athinoula A. Martinos Imaging Center at the McGovern Institute. Building on this knowledge, Gabrieli and others at the Martinos Imaging Center are working with Siemens to provide almost instantaneous feedback (in the form of a projected thermometer) to patients during functional magnetic resonance scans (fMRI). The feedback allows them to literally see if they are modulating their amygdalae in such a way as to reduce their depression. “We have shown that people can learn to master this. But the ability to do so depends on technical issues such as the speed at which feedback occurs,” says Gabrieli.

In order to provide nearly instantaneous feedback with MR—a technology that is known for its long scan times—Associate Director of the Martinos Center at McGovern Dr. Christina Triantafyllou worked closely with Siemens. “The idea was to get the data from the scanner to an external computer as quickly as possible, process it, and forward it to the patient as visual feedback,” explains Triantafyllou, who is also the Center's chief MRI physicist. “To achieve this, we developed a real-time data transfer system capable of keeping up with the substantial data stream generated from a 32-channel phased array head coil. We used a Siemens motion correction algorithm to reduce post-processing requirements and prototype real-time software from Siemens that corrects for geometric distortion and maximizes feedback times. We still have up to a five-second lag in the feedback loop, but that's because of the brain's own hemodynamic response to stimulus.” The technology could eventually also be used to empower patients to control addictive urges—including smoking and excessive eating, manage some forms of pain, and even improve concentration. Whatever its future applications, the technology that has made real-time fMRI patient feedback possible is a valuable tool on the road to development of next-generation MRI systems.

Better Images. Another major Siemens-MIT research effort is the development of an array of radio-frequency transmitters. “This project has enormous promise,” says Michael Hamm, who heads Siemens' MR research group in Boston. “Normally, you would transmit only one radio frequency into the body, switch it off and generate an image from the signals produced by a change in the spin of hydrogen nuclei, otherwise know as protons,” explains Hamm. “But with our array you can transmit multiple frequencies—also known as pulses—in parallel. The overlaying of parallel pulses can be used to shorten excitation duration, which may result in clearer, more reliable images at higher field strengths." The result is clearer images because the array can tailor the MR signal excitation profile and thus compensate for image inhomogeneities. “The array approach is one of the key elements of our collaboration with Siemens,” says Dr. Elfar Adalsteinsson, associate professor of health sciences and technology and associate professor of electrical engineering and computer science at MIT. “Its deployment to human imaging will be a breakthrough because its applications are so broad.“

An associated Siemens-MIT project that is also designed to accelerate scan times while producing higher-resolution images has focused on receiving MR signals as they return from the body. To image a patient's brain, for instance, the head is surrounded by antennas—also known as coil elements. Each antenna receives a signal from nearby tissues on its own channel. “The more coil elements—combined with newly-developed acquisition techniques—the higher the speed and resolution,” says Hamm, who points out that, “when imaging shallow areas, a 128 coil-element prototype system can accelerate signal processing by a factor of seven compared to a commercially-available 24 coil-element system.”

Improvements in resolution resulting from the use of more coil elements offer potential advantages for doctors and patients. Before brain surgery, for instance, surgeons want to know exactly where the patient's visual cortex is located in order to avoid any possible damage. “In this regard, we compared a twelve-channel head coil system with a 32,” says Triantafyllou. “We found that in order to determine the functional boundaries of the visual cortex, we needed five scans with a twelve-channel head coil, but only one scan with a 32 channel system. This allowed us to reduce the time for this study from 24 to four minutes.” (For more on the development of multi-channel systems, see Pictures of the Future, Fall 2005, Research Cooperation).

Pictures of Mental Illness. Thanks to the reduced scanning times and higher-resolution imaging that result from parallel transmission arrays and the ability to process information from more and more channels—not to mention more powerful MR scanners—researchers are beginning to use improved brain mapping to objectively identify the unique “signatures” of activity that characterize different psychological conditions. For instance, working with Siemens and Massachusetts General Hospital in Boston, Triantafyllou has used a specialized Siemens 32-channel coil for the study of children with dyslexia, autism and attention deficit hyperactivity disorder (ADHD). “The idea,” she explains, “is to document the development of the MR signature of each of these conditions over time with a view to improving diagnostic accuracy.”

Working along similar lines, Triantafyllou and Institute Director Gabrieli are studying MR signatures to determine whether it is possible to predict which dyslexic children will respond to treatment. “Some get much better, others don't improve. And conventional tests can't predict outcome,” says Gabrieli. “But with imaging we have been able to predict with 92 % accuracy which children would do a lot better. This is real evidence-based medicine; and it is based on the development of fMRI signatures that are strong predictors for improvement for certain conditions. We can digitally quantify the areas of the brain that are affected and compare results. All in all, this is a new field with huge social and healthcare cost implications.”

Rapid Localization of Strokes? As objective signatures for a range of conditions affecting the brain come into focus, they will be based on information pertaining to the precise anatomical distribution and quantification of physiological processes such as blood oxygenation—an area now being intensely studied by Dr. Adalsteinsson and his engineering PhD and M.D./ PhD students using Siemens scanners and specialized support.

“If we could quantify the rate of oxygen consumption in the brain, we could, for example, treat early-stage stroke much more effectively than is now possible,” says Adalsteinsson. “A radiologist would be able to see exactly where oxygen consumption was abnormal and recommend targeted treatment virtually on the spot.” The implications for this are huge. Each year, according to the American Heart Association's “Heart Disease and Stroke Statistics 2010,” nearly 800,000 Americans “experience a new or recurrent stroke… and on average, every 40 seconds, someone in the United States has a stroke.”

How close are we to seeing very early-stage stroke-related damage? “We are now at the point where we are building models of the rate at which the brain consumes oxygen in different areas and for different tasks,” says Adalsteinsson. “We need to quantify this in absolute units of energy consumption.” He cautions, however, that this is far more complex than it may seem: “Basically, all we have to go on in MRI is how much water is in a given location. That's not a lot of information. But there are factors in the local environment that affect that signal in subtle ways, such as its rate of decay, which in turn can be linked to levels of oxygenation through careful modeling. Imaging of appropriately-selected MRI signal sources in the brain, followed by estimation of their characteristic decay rates, can be used as a basis for a model of how the brain consumes oxygen. Doing this reliably requires very high quality MRI machines.”

Although water molecules constitute the major compound that responds to MR signals, there are other compounds that offer tantalizing prospects as new sources of diagnostic and research information. One such molecule, which, according to Adalsteinsson, offers a signature for “happy, healthy neurons,” is N-acetyl-aspartate (NAA). In patients with brain injury, stroke or Alzheimer's, NAA's signal will diminish over time, and could, for example, be used to detect the first signs of disease or track a patient's response to treatment.

“These are some of the questions we are exploring with Siemens and with our colleagues at Massachusetts General Hospital,” says Adalsteinsson. “In my opinion,” he adds, “what we are doing here is a prime example of what you can get out of a tight collaboration between academia and industry. As we turn the corner on new technologies and begin to deploy, we need the resources of industry, and that's where our synergy with Siemens comes in.“