50 years of MRI

Fifty years ago, chemist Paul Lauterbur of Stony Brook University published a paper announcing that he had found a way to use his nuclear magnetic resonance spectrometer to capture an image. As proof, he included a crosshatched picture of two capillary tubes of distilled water submerged in a tube of deuterated water. Though the difference between the liquids would have been invisible to other imaging methods, after some processing of the NMR spectra, the capillary tubes leaped out of a cross section (Nature 1973, DOI: 10.1038/242190a0).

Since that foundational paper was published, magnetic resonance imaging (MRI) has become a mature medical imaging technique—but it still depends on chemical principles to make tissues visible, and chemists are still finding new ways to enhance that visibility and make the most of MRI.

NMR principles

“MRI is the most complex diagnostic tool in radiology,” says Peter Seidensticker, a general radiologist and a vice president of global medical affairs radiology at Bayer. “To make it clinically feasible was, from the start, one of the key developments.”

As an imaging modality, MRI appeals to radiologists because of its high resolution in soft tissues and because it is minimally invasive. To create images, scanners image the protons in the body’s most abundant molecule, water, using the principles of nuclear magnetic resonance.

Simply put, NMR occurs when researchers provide a kick to aligned atomic nuclei in a magnetic field they have applied. The interaction between the nuclear spin—a magnetic property—the chemical environment, and the magnetic field produces an electrical signal that can be measured. By altering the sequence of radio-pulse kicks that they apply, chemists can further fine-tune the signal to get information about the environment that nuclei experience within or near the molecules they belong to.

By the mid-1960s, when commercial instruments became available, NMR was an established tool for analyzing molecular structures. But the technique was applied to samples in the lab, not patients in the clinic. Then in the 1970s, two chemists, Lauterbur and Raymond Damadian of the State University of New York–Brooklyn, separately suggested using the method to collect spatial medical information.

A historical invention or two

Damadian and Lauterbur both claimed to have invented MRI. The dispute became controversial, especially after the 2003 Nobel Prize in Physiology or Medicine recognized Lauterbur and physicist Sir Peter Mansfield but not Damadian. But Roger Turner, a curator at the Science History Institute, says the fight over priority doesn’t much concern historians. “What’s interesting about these controversies is that they reveal the multiple pathways into . . . developing a technology like this,” Turner says.

In his 1973 paper, Lauterbur describes how he obtained 2D data by adding a second magnetic field to alter the strength of the first field. This modification introduced position-specific changes to the signal, which could be interpreted to yield an image. Soon after, Lauterbur imaged a clam his daughter had found on the beach. Eventually, he collected the first 3D magnetic resonance image, capturing a coconut (Phys. Med. Biol. 1981, DOI: 10.1088/0031-9155/26/5/004).

Around the same time that Lauterbur was finding new uses for his spectrometer, Damadian observed that the NMR spectra of cancerous tissues were distinct from those of healthy tissues. He suggested using spectroscopy instead of visual examination to identify cancer in biopsies (Science 1971, DOI:10.1126/science.171.3976.1151; Proc. Natl. Acad. Sci. U.S.A. 1974, DOI: 10.1073/pnas.71.4.1471). Damadian patented the concept of NMR-based cancer detection and then in 1977 captured the first magnetic resonance image of a human (Physiol. Chem. Physics 1977, 9, 97).

Whereas Damadian started from a biological observation he hoped to apply medically, Lauterbur began with what Turner describes as “taking a very well-established machine and using it in ways . . . that the manufacturer wouldn’t recommend.”

MRI scanners were available in some hospitals by the mid-1980s, but in the decades since, the field has continued to evolve. One key area in which chemistry has contributed is in developing drugs to brighten images.

Medicinal chemistry for brighter scans

An experiment in Lauterbur’s first paper, which repeated the heavy-water image with a capillary tube of aqueous MnSO₄, demonstrates an important part of how magnetic resonance images are formed: differences in water protons’ chemical environment can cause dramatic changes in how they relax to the ground magnetic state after receiving a radio-pulse kick.

As MRI technology developed, scientists learned to use subtle differences in chemistry to spot the boundaries between tissues, but sometimes they needed more contrast. In the 1970s, researchers realized that ions such as Mn²⁺, Cu²⁺, Cr³⁺, and Fe³⁺, which are invisible to MRI, interact with nearby protons in a way that could be useful.

When a complex containing a paramagnetic ion tumbles in solution, it creates a small, fluctuating magnetic field that can shorten the relaxation time of water. That property, called relaxivity, depends on how the complex tumbles, how many water molecules solvate it, and how rapidly those water molecules exchange with the bulk solution.

Today, contrast-enhancing drugs are used to sharpen about half of MRI scans. Contrast agents that contain Gd³⁺ are the most widely used because of their high relaxivity. The most-used agents are carried in the blood and can spotlight problems like narrowed arteries or tumors by showing differences in the perfusion (or blood flow) or extracellular volume of different tissues. Peter Caravan, codirector of the Institute for Innovation in Imaging at Massachusetts General Hospital and a professor of radiology at Harvard Medical School, calls these agents effective but “in a nonspecific way.” Building in specificity is one property that researchers want to include in new contrast agents.

As well as designing more targeted molecules, chemists are developing alternative contrast agents to avoid rare but serious complications to which people with kidney disease are susceptible. There is also a growing recognition that gadolinium can accumulate in tissues after repeated scans. Although the long-term consequences of this accumulation are unclear, researchers wish to avoid any risks to patients.

“What you’re trying to find is an almost inert substance . . . that just goes through the body, is excreted in a predictable way, and doesn’t really interact with biomolecules,” says Seidensticker, whose colleagues at Bayer are testing a macrocyclic chelate called gadoquatrane as a new contrast agent in large clinical trials.

Seidensticker says that developing new agents is an exercise in trade-offs. For example, large, complex molecules that tumble more slowly tend to improve relaxivity, meaning doses can be lower—but they are also more viscous, a bad trait for an injected drug.

Some researchers are seeking to develop high-relaxivity complexes of other ions to find gadolinium-free contrast agents. Other classes of imaging agents under development go beyond bulk blood flow. For example, the drugs might target specific molecules within tissues or switch on in response to enzyme activity.

Other recently developed drugs used for enhancing MRI images belong to a class known as direct detection agents. These compounds introduce atoms—such as fluorine or, more recently, phosphorus—that are visible to MRI. Overlaying a fluorine scan with a proton scan can allow researchers to track labeled hot spots, such as cell therapies labeled with fluorine nanoparticles. But this technique requires a high local concentration of fluorine and faces increasing public concern about polyfluorinated organic compounds’ safety.

“In general, [contrast] agents do not provide direct therapeutic benefits, so the authorities will be quite strict in terms of what they would expect on the safety side,” Seidensticker says.

A fertile field

With an estimated 130 million scans conducted annually around the world, MRI is a mature technique, and some experts say there is little conversation today between MRI practitioners and NMR chemists. But according to biochemist Kevin Gardner, who directs the Structural Biology Initiative at the CUNY Graduate Center, advances in instrumentation or physics in one community still sometimes cross over to great effect.

One example involves magnetic resonance spectroscopy. Most MRI scans visualize the protons in water. But researchers can use MR spectroscopy to spot metabolic changes caused by diseases such as cancer by focusing on the protons in less-abundant molecules or on other nuclei altogether.

According to Brent Weinberg, a neuroradiologist at Emory University Hospital, the scope of proton MR spectroscopy remains limited by the low concentration of molecules other than water in the body. He estimates that he uses MR spectroscopy with just a small percentage of his patients, usually as a troubleshooting tool for tricky differential diagnoses. “It’s always been the dream that we could get this chemical composition of the entire area,” he says. “But the reality is we can see only a couple of the major metabolites.”

Researchers are working on ways to get more information from MR spectra—for instance, by automating analysis. But Weinberg says progress has been slow. Meanwhile, although MR spectroscopy of isotopes such as ¹³C, sodium, and phosphorus exists, Weinberg says it is not clinically useful because these atoms are vanishingly rare in the body compared with protons.

NMR chemists may have found a partial solution to the dim signals from heteronuclei. An emerging technique called hyperpolarization MRI borrows from dynamic nuclear polarization NMR to brighten the signals of nuclei by increasing their polarization. The resulting signal is short lived and may never see clinical use. But it lasts long enough that researchers can, for example, inject pyruvate with a hyperpolarized ¹³C into a person and watch how it is metabolized. “In MRI, we detect water because that’s the most abundant,” Caravan says. “But now, with hyperpolarization . . . you’ve got enough signal to image a specific molecule.”

Fifty years on, chemists are still contributing to MRI capabilities, and Gardner calls the conversation between the two disciplines a “very fertile field . . . for continuing development.” Who knows what chemists may contribute to MRI technology in the next 50 years to make even more hidden features of a living body visible.