Princeton Weekly Bulletin November 23, 1998

Improving NMR/MRI

	Warren Warren (l) and research associate Sangdoo Ahn with NMR spectrometer (photo: Susan Geller)

Chemistry professor's correction of 50-year-old theory results in clearer images

By Ken Howard

Little effects--such as the magnetic field generated by a lone hydrogen atom inside a large molecule--can have big implications. In the case of nuclear magnetic resonance spectroscopy, or NMR, this small magnetic effect has allowed scientists to analyze the structures of numerous compounds to a high degree of accuracy. It is among the most important analytical tools in the arsenal of the working chemist and biologist. Recently, however, chemistry professor Warren Warren detected a significant omission in the 50-year- old theory underpinning NMR. His discovery, reported in the July 10 issue of Science, has repercussions not only in the lab but also in the way doctors detect tumors.

Two mistaken assumptions

NMR spectroscopy, a technique for analyzing the molecular structure of compounds by measuring the energies of atomic nuclei (usually hydrogen), has served scientists well for over half a century. NMR worked well for Warren, too, until he began to experiment with unconventional ways of measuring differences in energy.

"A few years ago we did some straightforward NMR experiments and got seemingly impossible results. We then did hundreds of control experiments before we decided something was wrong with the theory behind NMR in solution. That was when the really hard part began, because the theory looked so good and seemed to have been so thoroughly tested over the decades."

Warren eventually discovered two mistaken assumptions: that magnetic fields of hydrogen atoms on distant molecules do not affect each other, and that at room temperature the initial state of the hydrogen atoms could be written in a particularly simple form. Removing either one of these assumptions by itself made no significant changes in the predictions, but correcting both of these assumptions at the same time yielded quite different results.

"This predicts that we can do entirely new pulse sequences," says Warren, referring to the pattern of radio frequency waves used to measure the energy differences between hydrogen atoms.

Tiny bar magnets

The magnetic fields on the hydrogen atoms make them behave like tiny bar magnets. In the absence of other influences, these atomic magnets line up in one of two directions. When an additional magnetic field is applied to a region of hydrogen atoms, as in NMR spectroscopy, nuclei aligned in the same direction as the applied magnetic field will decrease in energy level while nuclei aligned in the opposite direction will see their energy increased. Think of the energy one would use floating downstream versus energy needed to swim against the current. It is this difference in energy that is measured by the NMR technique.

Warren's corrections offer a more precise way of evaluating how hydrogen atoms are affected by their environment, thus yielding a better picture of that environment. The discovery will allow improvements in the lab (for example, instruments that use very powerful magnets can use this method to correct for field instabilities), but the more immediate impact will probably be with NMR's offspring, magnetic resonance imaging, or MRI. NMR and MRI are essentially the same technique, except MRI is used to analyze the contents of humans instead of what is in a test tube. (The name change is to alleviate patient fears associated with "nuclear" in NMR. MRI, explains Warren, "uses harmless radio waves instead of potentially dangerous high energy radiation.")

MRI

MRI, which allows physicians to view structural abnormalities and possible tumors inside the human body, works in the same way as NMR by detecting the presence of hydrogen atoms, this time specifically within water molecules. Humans are approximately 70 percent water, and MRI is able to differentiate fluid regions from solid areas such as bone, organs or tumors. An MRI machine does this by first creating a magnetic field around the patient, forcing energy changes within the body's own magnetic fields created by the hydrogen. The energy changes are measured by a sequence of radio frequency pulses sent through the patient. This information is then converted by computer into magnetic resonance images.

Warren's findings make it possible to improve the contrast in the resulting image. His team proved this approach on NMR machines using simple solutions, where signals predicted by the standard approach differed dramatically from those obtained using Warren's new pulse sequences. Researchers then verified the approach in animal studies at the University of Minnesota, where rats with brain tumors were scanned by MRI. The images of tumor regions using the new pulse sequences were enhanced compared to those taken with conventional MRI, and the new technique revealed structural features not visible in the conventional image.

Clearer pictures, better contrast

Because the new imaging technique requires no new hardware, only the rewriting of MRI pulse sequence programs, Warren's discovery may soon have applications in medical diagnostics. Validation studies have already been performed on healthy human volunteers at the University of Pennsylvania Medical Center using an MRI scanner. The results of the brain scans yielded "different kinds of MRI images than have ever been taken before, providing clearer pictures with better contrast," said Mitchell Schnall, associate professor of radiology at the University of Pennsylvania.

When applied to living tissue, the new technique allows MRIs to detect not only different water gradients, but also changes in oxygen concentrations within the body. This is significant in viewing tumors, because these cellular masses are often dead in the center and highly vascularized towards their exteriors; the new pulse sequences should be able to pick up this difference and thus produce better images of a tumor. "Things like iron and hemoglobin can make small magnetic fields," explains Schnall in reference to the oxygen-carrying components of blood. "By being able to see these better, we would expect to see tumors, because of their unique oxygenation state."

Plans are being made to begin clinical studies at the University of Pennsylvania to compare images of breast tissue taken using conventional versus the contrast-enhanced MRI to see if detection of tumors can be improved. Additional likely applications for the new pulse sequences include obtaining better contrast images for other types of cancers such as brain and prostate. The technique may also provide a new way to rate how advanced a tumor is by showing the amount of vascularization surrounding it, as aggressive tumors often require a large blood supply to feed their growth.

The corrections may also improve images obtained from functional MRI, a technique used literally to take pictures of the way people think. Says Warren, "Functional MRI is an extremely exciting frontier in nuclear science; you see what images in the brain are activated when someone does a task. The signal strength comes from changes in oxygen levels, so the new theory should improve the contrast of the images."