From those who describe their work as on the fringe of magnetic resonance (MR) research to neuroradiologists who use a range of brain imaging modalities, the capabilities continue to capture the scientific and medical minds at work in labs and clinics across the country. Surgeons, oncologists and patients are benefiting from scans that offer more data, increased insight and growing hope for people suffering from brain tumors, gene-linked diseases such as Alzheimers and Parkinsons disease, and autoimmune diseases such as multiple sclerosis (MS).
The views from the research lab and the clinic have never been better. The path to understanding, treating and eventually eliminating some neurological diseases through the assistance of increasingly high-quality images holds great promise. As research progresses, biologists are using imaging modalitieslong considered the clinical tools of radiologiststo examine the complexities of disease. The convergence of the two sciences undoubtedly will lead to opportunity for both groups.
On the clinical side, multi-detector CT is creating enthusiasm in the brain-imaging arena to view stroke damage by using contrast to obtain a map of brain perfusion. CT angiograms determine if there is an obviously occluded vessel. MR displays its technological prowess in imaging research settings with diffusion tensor imaging, which images white matter (the nerve tissue that contains large amounts of nerve fibers and, as a result, large amounts of the insulating material myelin) in the brain and is probably the quickest way to see damage from a stroke. While PET, the grandfather of the brain-imaging technologies, offers its impressive abilities to image tumors in cancer patients and nerve cell changes in the brains of people with Parkinsons Disease.
We have this big pallet of technologies, says Charles Strother, M.D., president of the American Society of Neuroradiology (ASNR of Oak Brook, Ill.). [The challenge is] how do we optimally apply these so theyre utilized in a manner that is good public policy, and how patients and physicians referring these patients understand what really is the best modality that gives them the best results.
Research being conducted by Eric Ahrens, Ph.D., assistant professor of biological sciences at Carnegie-Mellon University (Pittsburgh, Pa.) focuses on the development and application of magnetic resonance microscopy, extremely high-resolution MRI. The technology works off of similar principles as conventional MRI, but pushes the spatial resolution limitation.
What were able to do is achieve near cellular resolution using an MRI instrument, Ahrens says. The way we do this is by using an extremely high magnetic field and optimized detection circuits to get the ultimate sensitivity.
Using 11.7 tesla magnets in miniaturized MRI systems manufactured by Bruker BioSpin Corp. (Billerica, Mass.), Ahrens studies genetically altered mice to look at autoimmune disease, particularly in the central nervous system or the brain and spinal cord. He uses an animal disease model of multiple sclerosis called experimental allergic encephalomyelitis (EAE). The pathology of EAE mimics many aspects of human MS. In diagnosing and following the animal model disease, Ahrens is focusing on two aspects.
[We do] what I like to call virtual pathology, where on the basis of biophysical measurements that we can acquire using our magnetic resonance microscope, we try to determine what pathology is present in an EAE lesion at the cellular level, Ahrens says.
MRI has long been used to diagnose and follow MS. Typically, in an MR scan, an MS lesion looks like a light or dark spot in the brain. That appearance is consistent with many different types of underlying pathology, including the presence of edema or inflammatory cells or demyelination (damage of the myelin sheaths surrounding nerve fibers).
What wed like to do is be able to differentiate demyelination from inflammation and edema, Ahrens says. Demyelination is important because thats what causes most of the function deficits in MS patients, as well as the mouse models that we use. Were doing this by acquiring a series of biophysical measurements using our MRI machine.
The second aspect of Ahrens work is developing novel ways to track specific immune cell types in the body, particularly during autoimmune disease such as MS. The way were doing that is by labeling phenotypically defined sets of immune cells in vitro, reintroducing them into a sick animal, and then tracking their homing and migration in vivo, Ahrens says. This will allow us to visualize the earliest events of the onset of autoimmune disease and understand how the different subsets of immune cells are involved in the disease etiology. This will give us fundamental insights into the immunology of the disease.
In the future, people will want to use MRI to look at specific cells and molecules in the body to help diagnose disease like cancer and Alzheimers. But also people want
to be able to use MRI to monitor the efficacy of cell therapeutics and gene therapy, Ahrens says. Wed like to be able to monitor the biodistribution of specific molecules, enzymes or gene products. I believe thats the future of MRI.
These innovations will require the development of a new generation of contrast agents. These agents need to be selective for a certain cell type and they need to be able to be responsive to the presence of a gene product, for example. They need to be able to gain access intracellularly. The contrast agent renders these cells magnetically distinct, enabling them to be followed by the magnet.
I really think that MRI is on the verge of explosive expansion in the development of MRI reagents reminiscent of what was seen with fluorescent imaging technologies in the late 80s, Ahrens says.
The new cellular-molecular imaging capabilities can be performed on existing MRI machines. Ahrens sees that as a prime area for commercialization. I also think that MRI has become more and more a tool of the biologist doing basic science, Ahrens says. Since Im in a biology department, I try to show people how to use MRI to look at fundamental questions in biology in the study of development and disease, for example. I think youll see more and more research application for MRI. The pharmaceutical companies also are getting this idea because theyre investing lots of money in these small animal imaging systems. They understand that they can monitor the efficacy of therapeutic drugs or the effects of a compound non-invasively in the same animal over time, and that in the end can save them money and time.
UCLAs Paul Thompson, Ph.D., has used MRI to track the loss of brain tissue in cases of Alzheimers disease, dementia, and other degenerative diseases.
Maps of change
At the University of California at Los Angeles (UCLA), Paul Thompson, Ph.D., assistant professor of neurology, uses MRI in his research to reconstruct a time-lapse visualization of the spread of Alzheimers disease. The idea is to scan patients longitudinally, having a scan every three months or so, then plot the time course, Thompson says.
The information could be useful for early diagnosis because it shows early changes. It also could be used in drug trials to determine if the drug is saving brain tissue or slowing the rate of loss. Even though with MRI you get a static picture of the brain, we think if you can do this longitudinally, even the tiniest changes, maybe 1 percent slowing of the disease, could be detected, Thompson says.
The approach also is applicable to dementia and other degenerative or progressive diseases. Normal, healthy people lose less than 1 percent of brain tissue per year. People who are beginning to be at risk for Alzheimers lose just above 1 percent of brain tissue per year. Alzheimers patients lose approximately 5 percent of brain tissue per year.
In dementia in particular, changes on images actually precede the overt symptoms of the disease, so you want to give the medication as soon as possible. Its known in Alzheimers that certain parts of the brain degenerate three to four years before the symptoms actually begin, Thompson says. With MRI, even though its conventionally used as a single-time-point modality, people are using it increasingly to follow 0patients over time. You can get a lot more information that way because changes are happening that are very tiny, maybe less than 1 percent per year. Such small changes are typical of healthy aging or the early signs of Alzheimers, and you could really pick that up from a sequence of MRI scans, using the patient as his own baseline.
The information is fed into a Silicon Graphics Inc. (SGI of Mountain View, Calif.) supercomputer. Subtle changes that could be missed by the trained eye of a radiologist are detected by the computer, which Thompson says is ideal for the task.
We use the SGI computing technology to compute very detailed maps of brain change, Thompson says. The way it works with computers is we actually elastically morph the earlier [baseline] scan onto the later one. This is a quantitative way of dealing with brain images where you take a baseline scan from a patient and ask at what rate the disease is accelerating.
Thompson makes a color-coded picture exactly where the brain is losing tissue or, in the case of brain tumors, where it is gaining tissue rapidly. The technology presents a computer-assisted image, providing more information than from the scanner itself. Its sort of distilling the finest details out of the scans, especially if theyre subtle and color coding the regions that change the most, Thompson says. The tiny loss of 1 percent or so is undetectable without the assistance of a computer, according to Thompson.
As more powerful methods for detection appear on the scene, earlier diagnosis becomes closer to reality. Clinical trials are now looking at people at risk and giving the medication early to prevent or delay onset of symptoms. Theres a lot of evidence that will work, Thompson says. It wont prevent dementia, but if there is a change, it will delay the onset of symptoms maybe two or three years. Thats how its being used for these at-risk people. People at risk include those with a brother, sister or parent who had early dementia or people with the APOE4 gene (estimated to be a quarter of the population), a concrete genetic marker linked to early onset Alzheimers.
Robert Moore, M.D., Ph.D., The Love Family professor, professor of neurology and neuroscience and co-director of the National Parkinson Foundation Center of Excellence at the University of Pittsburgh (Pittsburgh, Pa.) uses PET in his work with Parkinsons disease. It is a technology that allows you to image specific groups of nerve cells in the brain, and in Parkinsons disease, the principle important group of nerve cells that is affected produce the chemical dopamine, Moore says. There are PET techniques that allow you to visualize those nerve cells to study exactly which of the nerve cells are affected, the way in which theyre affected in the areas they innervate and in looking at the progression of the disease and hopefully in the future looking at the effects of treatments on the progression of the disease.
Papers have been published in the last year on the effects of neuroprotective agents, which would slow the progression of disease in Parkinsons patients. PET is one of the best ways for measuring the progression of the disease, Moore says. And that has been employed.
A PET scan to image the brain for Parkinsons takes between an hour and an hour and a half. Patients do not experience discomfort. The technique requires the patient lie still in a recumbent position. The patients head is put in a plastic holder that keeps it still, which Moore says most people find quite comfortable. The radiopharmaceutical used is fluorodopa. Moore uses a Siemens Medical Solutions USA Inc. (Malvern, Pa.) ECAT Exact 966 scanner.
Functional MRI (fMRI)
Although functional MRI can be used to look at the function of the brain by looking at blood flow, and it looks at all kinds of nerve cells, it does not specifically allow you to look at the ones that are selectively affected in Parkinsons and other similar diseases. It has, however, found amazing utility for other conditions.
Surgeons are beginning to use fMRI to plan very complex surgeries where it is necessary to know the function of different brain tissues. Its mainly a research technique, UCLAs Thompson says. It tells you which part of the brain is active when you do a particular task, and thats something that is really quite phenomenal for people who are interested in brain function and how the brain is organized.
The functional imaging scan is more difficult to do and takes approximately 40 minutes compared to eight minutes for standard MRI scans. The fMRI scan provides a picture of which parts of the brain are active during certain performed tasks. It enables a surgeon to know the eloquent, motor or sensory areas of the brain to avoid during a complex tumor or venous malformation surgery. The area is mapped and avoided.
There arent many completely set paradigms that can be done in an average clinical setting where everyone is doing the same type of test, enabling comparison from one site to another, however.
Diffusion tensor imaging (DTI)
Research labs are providing better views of stroke victims through diffusion tensor imaging. Ana Solodkin, Ph.D., research assistant and associate professor at the University of Chicago uses the technology to assess the damage produced by stroke, focusing on motor recovery. Solodkin looks at patients one month after their stroke, following them and measuring their motor behavior.
At the same time we are doing this, we do functional imaging of the motor behavior as well, and with that we can assess the changes in the brain and activation of the brain through recovery, Solodkin says.
Solodkin and her colleagues do network analysis, considering how areas of the brain are relating to each other. Using DTI, they assess the connectivity of an area and create a functional map. We take the areas that are active during bilateral movement, for instance, and map the connection from point A [the primary motor area] to point B [the secondary motor area], Solodkin says.
With a stroke patient, you have a patient who cannot move a hand, and if you ask that patient to move that particular hand, there is no way that can happen, but then if you ask a patient to move both hands at the same time, they can move both hands, Solodkin says. So what we have studied with our networks is the relationship with one side of the brain to the other side.
They also assess normal perfusion imaging to help determine if the central nervous system as a result of a stroke is changing its metabolic requirements.
Diffusion tensor imaging is in development and not yet commercially available. Solodkin uses a GE Medical Systems (GEMS of Waukesha, Wis.) three-tesla Signa functional MRI system.
Susumu Mori, Ph.D., associate professor of radiology at The Johns Hopkins University (Baltimore, Md.) uses diffusion tensor imaging to view white tracts in the brain. Some people may believe in [diffusion tensor imagings] ability to study brain malformation or tumor, Mori says. One thing for sure is that this imaging can delineate white matter structure much, much better than the conventional imaging. So any anatomical change related to white matter structure can be better delineated by this technique.
Mori and his colleagues have shown DTI gives good anatomical information of cerebral palsy. Although conventional imaging can provide easy diagnosis, it cannot clearly show which white matter tracts are severely damaged and which white matter tracts are spared.
Research at UCLA used MRI to reconstruct a time-lapse visualization of the spread of Alzheimers disease.
In clinical practice
The East Portland Imaging Center (EPIC of Portland, Ore.) recently installed a new Intera 3 tesla MR scanner by Philips Medical Systems (Bothell, Wash.). The center does general diagnostic brain imaging using mainly MR and some CT scanning, providing about 105 brain MRIs per month (and 60 brain CT studies per month). Installing the 3T MR scanner enabled the facility to do MR spectroscopy to evaluate relative metabolite concentration to the volume of brain tissue. Although some facilities use 1.5 tesla magnets for spectroscopy, Simon Roman-Goldstein, M.D., a neuroradiologist at the center, says they opted for the higher magnet strength.
I think all the spectra are better on a three-tesla magnet, Roman-Goldstein says. With MR spectroscopy, two of the major uses are to help characterize an abnormality that we see, so if we see an area that shows abnormal signal or abnormal enhancement, we can get a spectra and help characterize what type of lesion is causing that abnormality. Or alternatively, you can do spectra of a large body of brain tissue and then try to get an idea of the extent of the abnormality.
Looking at three major metabolites, including markers of neuronal cells, membrane turnover and energy and acidity helps in conjunction with the signal intensity changes on standard MR imaging. They enhance and characterize masses.
One of the problems now is most spectroscopy is done over a relatively large volume of brain tissue, Roman-Goldstein says. You can average in normal tissue with abnormal and not really get a completely characteristic spectra. As the techniques improve and you get smaller volumes, I think youll be able to get spectra that really only has areas of abnormal brain tissue, and it may help in better characterization.
CT continues to move forward in brain imaging work. The multi-detector scanners allow for much faster imaging. They show their real speed advantage in an acute stroke setting (CT is usually more available than MR in emergency rooms) to see if there is any hemorrhage or an obvious large stroke. Debates remain whether MR can completely exclude an acute hemorrhage.
Often in a large stroke, you dont see anything in the first few hours with a CT scan, Roman-Goldstein says. Then with contrast and the newer, faster scanners, they can get a sort of map of the brain perfusion by seeing where the contrast goes, as well as do a CT angiogram to see if theres an obviously occluded vessel.
Prior to the 16-slice CT, getting a perfusion map was harder with CT than MR. MR diffusion imaging is probably the quickest, earliest way to see damaged brain from a stroke. Youll see signal intensity change before youll see signal intensity changes on the standard T-2 or T-1 images, Roman-Goldstein says. And youll see those signal intensity changes before youll see signal density changes on the CT scanning.
The center also has a PET scanner on site. The main use is with fluorodeoxyglucose (FDG) to determine increased glucose utilization in tumors vs. areas of diminished glucose in the brain from radiation necrosis, which can look quite similar on anatomic and CT imaging.
If they were to take it [the lesion] out and do an en bloc section of the whole lesion, there are often areas of both recurrent tumor and radiation necrosis, Roman-Goldstein says. Sometimes you have to coordinate the MR imaging, the spectroscopic and the PET scanning to try to get a better idea [of what is present].