Amy Taylor (not her real name), a 30-something mother of two with a palm-size opening in her skull, tries to muster a smile as she sits up in a bed in the epilepsy monitoring unit at Yale-New Haven Hospital. A grid of depth electrodes penetrates the lumpy wet crown of her exposed neocortex. More than a dozen wires lead to a computer waiting to register the electrical firing patterns of the next in what has become a regular but unpredictable series of epileptic seizures.
The medications she has been taking for more than a decade no longer control those seizures, which are the result of temporal lobe epilepsy. She cannot work or drive, and rarely leaves home anymore. She will remain in the unit for up to 10 days, until enough data have been collected to localize the source of the malfunction in her brain. Electrical stimulation will also be fed into the electrodes while she undergoes a battery of tests to determine the precise location of critical language, movement and sensory areas.
Despite her present misery, she is fortunate. Hers is the only type of chronic brain disease that can be cured with surgery. But only the results of the tests will reveal whether she is a candidate for treatment.
Two to three patients each month undergo intracranial surgery for the type of testing Taylor underwent at Yale as a first step toward surgery to cure her epilepsy. The Epilepsy Education Association estimates that 20,000 Americans with intractable seizures could be cured with surgery, yet each year only 500 undergo the procedure. Fear and expense keep many patients away. Altogether, the procedure can cost more than $120,000. Using new imaging technologies, Yale scientists hope to make the choice of surgery far less frightening and more economical. A multifaceted team of engineers, linguists, physicists, neurologists, neuroscientists, applied mathematicians, computer scientists and neurosurgeons is refining methods for applying emerging magnetic resonance imaging (MRI) and spectroscopy (MRS) technologies to epilepsy. If they succeed, sometime within the next five years they will be able to locate the seizure focal point in a patient’s brain and map out areas of brain function in less than three hours—a tiny fraction of the 10 days Taylor spent under observation—and without opening the skull. Just eliminating that part of the surgical process will reduce both the total cost of epilepsy surgery by up to 40 percent and a source of enormous fear for patients.
The Yale investigators’ ambitions run well beyond that. They hope that sometime within the coming decade temporal lobe epilepsy will be curable without any surgery at all. According to Taylor’s neurosurgeon, Dennis D. Spencer, M.D., HS ’76, “It all hinges on our being able to use magnetic resonance imaging machinery to measure what’s happening in the human brain.”
New windows, new insights
Spencer, director of the Epilepsy Surgery Program at Yale and chair of the Department of Neurosurgery, already relies on MR studies to help guide him when he operates. The possibilities for developing treatment methods that, according to Spencer, “will put me out of business” depend on studies under way at Yale’s Magnetic Resonance Research Center (MRRC). Already one of the world’s leading scientific research programs of its kind, the MRRC received a major boost last winter with the opening of its new home in the 457,000-square-foot Anlyan Center for Medical Research and Education.
The MRRC’s 50-member staff, along with six multiton magnets supported by banks of computers, relocated from the Fitkin Memorial Pavilion basement to more than double the space in a two-story facility in the Anlyan Center (See The Big Move, Winter 2003). On the upper floor of the new facility, faculty and administrative offices surround a large, light-filled open work space with computer workstations at its center. The facility has 40 networked computers dedicated to analyzing data produced by its magnets. At the computers, students and other investigators manipulate images of organs and graphic displays of data and develop algorithms to model the behavior of parts of atoms within cells.
For the faculty and staff, the new quarters represent the increasing importance of imaging technology to biomedical science and health care—and a big boost for morale. Director Douglas L. Rothman, Ph.D., professor of diagnostic radiology, notes that this is the first time he has had a window in his office since he came to Yale as a graduate student in 1985. More important than what he can see out the window is what the new facility will enable imaging scientists to see at the molecular level. “This,” he says, “is one of the best facilities in the world now.”
On the floor below, the MRRC houses eight magnets, including two newly purchased systems. The MRRC maintains three animal and tissue research systems, including a soon-to-be-installed 11.75-tesla animal magnet, able to measure changes in animal metabolism at the molecular level. (A tesla, named for radio-engineering pioneer Nikola Tesla, is a measurement of the strength of a magnet’s field.) The center also houses three human systems, including a new 4-tesla functional MRI (fMRI) system in the W.M. Keck High Field Magnetic Resonance Laboratory that can pinpoint functional activity in an area as small as 500 micrometers across. That is half a millimeter, about the size of the fundamental information processing units in the human brain (often referred to as cortical columns).
The state-of-the-art, $3.7 million human magnet was paid for in part through a $1 million grant from the W.M. Keck Foundation. There are also empty bays waiting for eventual installation of two additional magnets. All of the MRRC bays are encased within 11 inches—nearly 2 million pounds—of steel and copper to shield out even the most minute radio signals, which could wash out reception by the ultrasensitive magnets, and to contain their magnetic fields.
The new magnets have already dramatically improved what investigators will be able to visualize. Studies of the central nervous system will benefit particularly from the magnets’ higher sensitivity. “They allow us to look at fundamental neuronal processing units, which we could not do with our previous equipment,” says Rothman. For instance, tissue that causes epileptic seizures can now be studied to determine precisely how metabolic processes critical to normal neuronal activity have malfunctioned, information crucial to the development of new treatment methods.
MRI and MRS studies in many fields at Yale, including reading and dyslexia, substance abuse, diabetes, mental illness and cardiovascular medicine, are providing methods for tracking down disease mechanisms and leading to novel forms of treatment. Basic science research at the MRRC is contributing to a new understanding of fundamental physiological mechanisms such as how energy is metabolized in the brain, heart and muscles. MR physicists at the center have also expanded the understanding of the underlying physical principles of MRI and MRS.
The MRRC is both a core center serving the research needs of the entire university and the primary research space for seven Yale faculty. These faculty have primary or secondary appointments in the Section of Biomedical Imaging in the Department of Diagnostic Radiology, which was formed to synergize the expertise of the imaging research groups at the School of Medicine. About 50 separate grants, totaling nearly $35 million and representing 50 faculty members in 13 different departments, rely upon the MRRC. The center itself is supported by fees from those grants and $7 million in direct research grants to faculty in the Section of Biomedical Imaging. The section is linked to the Department of Biomedical Engineering, which is part of both the medical school and the Faculty of Engineering on the main campus. In 2002, the new National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health awarded its first-ever grant to the section, $1.3 million to develop technology to map neocortical epilepsy.
Faculty and staff at the MRRC work on their own research advancing imaging science and also devise new ways to apply imaging technologies to biomedical problems. “We’re developing state-of-the-art magnetic resonance techniques,” says R. Todd Constable, Ph.D., associate professor of diagnostic radiology and neurosurgery and director of the fMRI group at the MRRC, “and as a function of that we’re able to provide a state-of-the-art resource for the whole university.”
MRI is a complex technological feat, combining physics, mathematics, computer sciences and biomedical science. The hauntingly clear MRI pictures result from the differing radio signals given off by protons in the atoms of tissues exposed to the intense magnetic fields. Those signals must then be interpreted by computers and translated into graphic imagery. The MR data require extensive manipulation, particularly in complex experiments, to be translated into readable images. Rothman says, “Studies rarely work straight out of the can.”
Constable and his 20-member team provide imaging support for more than 30 faculty members directing major grant-supported investigations utilizing fMRI within and outside the university. These include efforts to improve the understanding of autism with Robert Schultz, Ph.D., in the Child Study Center; studies of the role of the frontal lobe in working memory and executive processing with neurobiologist Patricia S. Goldman-Rakic, Ph.D.; mapping of memory processing with psychologist Marcia K. Johnson, Ph.D.; and experiments aimed at finding better treatments for gambling disorders with Marc N. Potenza, Ph.D. ’93, M.D. ’94, and for schizophrenia with Bruce Wexler, M.D., FW ’77, both in the Department of Psychiatry.
One of the most active programs utilizing MRRC resources is the Yale Center for the Study of Learning and Attention, which has pioneered the study of pathways the brain uses for reading. Co-director Sally E. Shaywitz, M.D., also directs the Connecticut Longitudinal Study, which is investigating the development of reading skills in children from ages 7 to 18, and in adults, and is the largest MRI study ever undertaken in children. With the resources of the MRRC, Shaywitz and her husband and research partner, Bennett A. Shaywitz, M.D., have mapped the neural systems responsible for reading. They have also begun to explore the impact on those neuronal pathways when educational techniques are used as interventions to help children with dyslexia overcome their reading challenges. “We can see how the brain is responding,” Bennett Shaywitz says of those education interventions. “If we see effects on the brain, we’ll know we’re on the right track.”
The Shaywitzes have begun to explore the use of MRS, a technique for measuring metabolic rates within tissue. The technique can measure the increased presence of a harmless biochemical tracer linked to glucose to show where increased or decreased metabolic activity, associated with tissue function, is occurring. The Shaywitzes will use MRS to search along the neural pathway for reading to find where disruptions underlying learning disorders may be taking place. “We are addressing questions at a finer and finer level,” says Bennett Shaywitz. By using the brain imaging technology they hope eventually to diagnose learning disorders early enough to prevent lifelong difficulties.
Diabetes and brain energetics
Robert G. Shulman, Ph.D., founder and for many years director of research in the MRRC and now Sterling Professor Emeritus of Molecular Biophysics and Biochemistry, oversaw installation of the first human-sized research magnet in Fitkin in 1986. He pioneered many of the MR techniques now used in biomedical research around the world. His work demonstrated the validity of MRS studies for tracing metabolism in tissue and then showed that the technology could be used to explore the role of metabolic pathways in humans. Many clinical advances followed those early studies.
In collaboration with professor of medicine and Howard Hughes Medical Institute investigator Gerald I. Shulman, M.D., Ph.D. (no relation), Shulman showed that a defect in muscle storage of glycogen was responsible for the impairment exhibited by type 2 diabetes. That was in 1990. The following year, Robert Shulman, Rothman and James W. Prichard, M.D., a former professor of neurology who is now retired, were the first to demonstrate the effectiveness of MR to follow brain functional activity, now a major application of the technology. Shulman also directed studies that proved the feasibility of using MR to study regional metabolic functioning in the brain.
Much about the brain remains a mystery, including how it utilizes energy to carry out its most complex functions. Robert Shulman, Rothman, and Fahmeed Hyder, Ph.D., an associate professor of diagnostic radiology in the MRRC, have made a number of recent findings using MRS that have sparked an emerging re-evaluation of the nature of brain function. These findings were made possible by the development of MRS methods to simultaneously image brain energy consumption and the release of glutamate and GABA, the primary excitatory and inhibitory brain neurotransmitters, respectively. Their studies have shown that even at rest the brain uses 80 percent of its energy to support neuronal firing and neurotransmission, which are the bases of brain function. Previously it was felt that there was little neuronal activity in the resting brain.
Understanding the role of this enormous underlying activity has implications for brain function studies of all sorts. “The baseline is not negligible,” says Robert Shulman. Rothman adds, “Our brains are always ‘on,’ despite the high energy cost.” Understanding why this resting neuronal activity is necessary could help explain the complexity of higher-order brain functions, such as learning and memory, and potentially open up new pathways for understanding disruptions in those functions. “These results show that the popular analogy that the brain acts like a computer, which only accesses its processing power when it is called on to perform a task, is not correct,” says Rothman. “A new view of the cognition will need to account for the data.”
Back to the OR
Novel understanding of brain metabolism made possible by Rothman, Robert Shulman and others has radically altered neurosurgeon Spencer’s outlook on epilepsy and its potential treatment. Epileptic seizures are a state of electrical hyperexcitation that starts from a single site in the brain and then spreads swiftly. Studies done at the MRRC demonstrate what Spencer terms a paradox. The focal point triggering the seizure is metabolically depressed relative to its surroundings. Spencer, Rothman and Ognen A.C. Petroff, M.D., HS ’82, associate professor of neurology, have been carrying out MRS studies to explain that paradox. MR analysis of tissue removed from patients with epilepsy showed that it was not processing glutamate appropriately. Instead, due to impairments in the cellular energy supply needed to keep extracellular glutamate at a safe, low level, the glutamate outside the cells was “backing up,” eventually setting off an electrical fluctuation that spreads quickly into a seizure. This and related findings may provide a key to treatment advances.
“This opened an important little door where we’ve not had many thoughts about new therapies,” says Spencer. “If it is an energy problem, then this is perhaps reversible.” Delivering medication to the site of the brain malfunction could reregulate the brain’s metabolism, much like pace-makers now do for patients with heart disease, to prevent seizures. Utilizing new technology in the MRRC, he says, “we can now create animal models to think about therapies. The underlying cause of epilepsy may be reversible, but the only way to know is by imaging before and after surgery.”
His work with the MRRC has already paved the way for better and potentially less-invasive treatments for epilepsy patients. In association with Spencer, the MRRC has developed new ways of bringing imaging techniques directly into the operating room. This summer, Spencer will begin operating on patients, utilizing computer-guided equipment developed by the Yale team in conjunction with the firm BrainLAB. He will operate while wearing goggles with a built-in monitor that gives him access to a virtual three-dimensional display of MR imagery from inside patients’ brains. The images, linked to the placement of the scalpel, will show him precisely where the cutting blade is located beneath the surface of the brain and what functional areas it is near. That should help reduce damage to surrounding tissue. Already he utilizes a portable version of the BrainLAB to give him visual information about the patient’s brain as he operates. “It looks just like a tricorder,” he says, referring to the medical device used in the futuristic television and movie series Star Trek.
James S. Duncan, Ph.D., vice chair of diagnostic radiology and director of the Image Processing and Analysis Group, is helping to develop even more futuristic treatment methods. His group developed the software used to integrate spatial, functional and metabolic images recorded in MR studies into the real-time spatial coordinates of the patient’s brain as Spencer operates. He is also attempting to develop a model that simulates the natural deformation of the brain that takes place during open-skull surgery. “If you are off by even less than a centimeter, you can hit critical areas,” Duncan says. In the coming years, Spencer will use Duncan’s models of brain deformation while he operates. The two expect that, within the next five years, MRI and MRS testing to identify the source of seizures and map the functional areas of the brain in combination with Duncan’s graphic models will eliminate the need for the costly, arduous and potentially hazardous intracranial surgery that Taylor underwent.
Spencer credits the scientists at MRRC with changing his view of epilepsy. “Instead of thinking of epilepsy as an electrical problem,” he says, “I now think of it as a metabolic disease. They are much more than MR physicists. They make you think about disease-related problems in a different way.” YM