Paul Lauterbur and the Invention of MRI

by M. Joan Dawson

  The technique that I have developed is, to my knowledge, completely new. It is a new form of microscopy, permitting the observation of the distributions of atomic nuclei and unpaired electrons in objects, from rocks to organisms. It can supplement X-ray and ultrasonic imaging of whole organisms and structures. Recent experiments, not quite ready for publication, have revealed the arrangement of the soft tissues of a clam inside his closed shell, and their isotopic composition, as well as demonstrating that features of tissues characteristic of malignant tumors can be detected inside an intact organism. Flow rates in hidden blood vessels, the concentration of sodium in and around cells, and the locations, within intact plants and animals, of the unpaired electrons generated during photosynthesis and various enzymatic processes may all be measured by variations of the basic technique. These possibilities, and many more, are quickly grasped and enthusiastically elaborated upon by most of the chemists, physicists, biologists, and physicians with whom I have discussed my results, and I have no doubt that your readers would respond in a similar fashion.24

  This time the paper was sent to an anonymous referee, who found that Paul

  . . . describes an ingenious nmr experiment. In itself the example he describes is a trivial one. The initial impression is that it is difficult to envisage that the technique would find widespread use in magnetic resonance investigations. However the author does state that information on chemical and isotopic composition and relaxation data can be derived. If this is correct, and being aware of Professor Lauterbur’s work and reputation, this statement would be accepted, the technique has a high probability of proving a useful one. . . . Certainly, if I were not aware of Professor Lauterbur’s eminent reputation I would not recommend acceptance without further evidence.25

  The revised paper, identical to the original except for an added final sentence—“Zeugmatographic techniques should find many useful applications in studies of the internal structures, states, and compositions of microscopic and macroscopic objects”—was accepted, but without the word “macroscopic.”

  There is a sequel. In fact, there are two sequels. The first was that thirty years later, Nature published full-page ads, patting itself on the back for being the first to press with the development of MRI (“That’s sweet revenge,” Paul noted). Shortly thereafter, the editors of Nature published a book, A Century of Nature: Twenty-One Discoveries That Changed Science and the World.26 Among the twenty-one most influential papers of the twentieth century was Paul’s.

  Sunshine of Progress

  People often ask how big the laboratory was in which this great work was accomplished. The answer is astonishing; Paul worked alone. He had no funding for his research in 1971–73, and therefore no group of students to work with him. “The equipment was being used for real chemical work during the day” Paul said, “so in the evenings, I’d put bits of plants, earthworms, clams, fruit, you name it, into the spectrometer and just look at the MRI signals.”27

  Within his own department, Paul’s work on MRI was met largely with puzzlement. Paul gave his first talks on MRI at his own university and at the nearby Brookhaven National Laboratory. Few of his fellow faculty really understood what he was up to, and at least some objected to the use of departmental resources for a project that, whatever it was, was certainly not chemistry. In this atmosphere, Paul was and ever remained grateful for the support and encouragement he received from his department head, Francis Bonner, who asserted, “What a chemist does is chemistry.”

  Figure 6.4

  Who put this in my wallet?

  There were exceptions to the general bewilderment. In general, the inorganic chemists were quicker to appreciate Paul’s new line of work than were others. Harold Friedman, a physical chemist interested in the properties of fluids and solutions, quickly caught on to the concept and its importance, as did assistant professor Charles Springer, who became a distinguished researcher in MRI and biological spectroscopy. Charlie tells of working late and finding that Paul was still in his office around midnight. Charlie made a point of stopping by, in part to show the senior faculty member that he, too, had his own nose to the grindstone.

  Paul often took the children to the beach on Long Island Sound for nice summer Saturdays. On one of these outings he sent his daughter, Sharyn, and son, Dan, out to look for the telltale water spouts and bring back clams. Then nine, Sharyn remembered it as a big moment of triumph in her sibling rivalry with her brother. Dan came back with adult clams. She went for the cute babies, the little ones. “It’s a guy thing, you go for the big ones. But in this case, it’s small size that mattered. Dad sifted through and chose my clam.” As the news release for the Kettering Prize subsequently noted, “That clam became the first living thing imaged by MRI.”28 These tiny Long Island clams were among the few living things big enough to image yet small enough to fit into the magnet Paul had available, 4 mm diameter of usable space. The clam image looked “kind of like a clam if you had a little faith,” Paul said.

  Progress was quick. Nature received Paul’s original paper showing two-dimensional MRI in October 1972. Paul followed this with a talk at Argonne National Laboratory in May 1973, in which he demonstrated that three-dimensional images could be produced and that isotopic exchange could be imaged and that water diffusion could be imaged and measured by MRI.29 The physical example he used in this talk was the two-capillary setup shown in the Nature paper. The biological example was a parsley stem. The talk was scheduled late on the last afternoon of the meeting, and hardly anyone showed up. He’d had no impact, so he had to try again. This kind of thing went on all his life. There are people walking around today thinking that they were the first to do these things, and they weren’t even born in 1972!

  By 1973, Paul’s technique had evolved. The data were now recorded using a paper chart recorder and digitized by measuring the heights of the peaks using a plastic ruler. Paul punched the data onto computer cards and fed them to a computer that had 16K of memory. Each step of the calculation had to be done with a new set of cards. The computer kept giving cryptic, indecipherable error messages, and someone had torn out the relevant pages of the book on error messages. That year Paul gave an impromptu talk at the Experimental NMR Conference (ENC) showing images of the two capillary tubes, the clam, and an earthworm, all accomplished with a standard NMR magnet.30

  In his 1974 talk at the same conference, Paul had graduated to imaging twenty-one capillary tubes and a section through the thorax of a mouse. These required both a magnet with a larger bore, one he obtained as a castoff from a company that was shutting down, and more complex data analysis. The data processing was not yet under complete control—a large bright artifact appeared in the center of the image that was in no way related to mouse anatomy. But this time, the magnitude of Paul’s discovery exploded into view.

  The World Responds

  The reception of Paul’s new ideas was much better in the UK and in Europe than it was in the United States. Paul felt that the British had much more patience with the basic science developments than the Americans did. He wrote,

  During the mid-1970s, however, I found that thoughtful discussions of the implications of the idea were seldom possible with physicians in the U.S., who were put off by the unfamiliar nature of the concepts and technology, and the obvious need for at least several more years of development before human diagnostic tests would be possible. During a number of trips to the U.K., I found that many physicists, medical physicists and physicians there found the novelty fascinating rather than disturbing, and expected that there would, of course, be a period of refinement of the ideas and development of apparatus before clinical trials could begin. Representatives of U.K. government agencies also convened a meeting with interested scientists and physicians to formulate a coherent policy for carrying the work forward, and aggressively pursued the patenting of all developments in U.K. Universities.31

  On the commercial side, Paul complained that U.S. company representatives gave him elementary
lectures about the cost of capital, return-on-investment curves, and the general impossibility of ever making a profit on anything really new that required R&D expenditures. One remarked, “Perhaps we’ll take a look at it after NIH puts in a few million to make it work.” In summary, Paul said, “European scientists, physicians, governments and industries moved more confidently and thoughtfully into this new area than did their American counterparts. When American companies did move in, they often bought expertise from the U.K.”

  In the UK especially, and Europe in general, people were not only accepting but excited about the possibilities. It is no accident that much of the early technological development of MRI was done in Britain, where scientists did not face the “results now” constraints of the Americans.

  Where Is the Money?

  Paul understood the importance of improving imaging and image-processing methods if they were to be diagnostically useful. And he knew he would have to demonstrate diagnostic usefulness to gain attention from the medical world. This is a very tall order for one small laboratory. The earliest work on MRI received no direct funding from the government or any other source. Paul did receive a couple of thousand dollars from the university, with which he could carry out the experiments necessary to apply for a grant from the National Cancer Institute. The only animal he referred to using in that application was a clam. People on the review committee at first decided that the stuff was absolutely nuts. But, knowing that Paul was not nuts, they took another look at the application. Before rendering their final judgment, they said, “We still can’t really understand it and it still looks crazy but if you read it carefully, everything looks as if it works out”—and approved it. With this funding Paul was able to hire a postdoc and graduate students and to start acquiring some equipment. He invented MRI in 1971 and received no real financial help to develop it until 1974. This was to be the pattern with every new idea he produced.

  Asked by British scientist Austin Elliott about funding, Paul told him,

  For the first work I did it myself on an existing [NMR] machine in the department, so just my time was involved. Later the first funding was through a mechanism that the NIH had at the time in which they would give a certain small fraction of the institution’s NIH grant to the institution to spread around in whatever way it thought was useful for the early stages of research. So I got probably over $1000 or $2000 or something from that for very minor things—it was all very cheap at the beginning. All the early tests and the mathematical ideas were done on paper with square grids in which I carried out the mathematics by hand—which again was just my time. There was no funding involved.32

  Meanwhile at Home

  For all his outward success, Paul’s life at this time was difficult because of his wife’s problems. The Chemistry Department’s administrative assistant told me how Paul used to go around with “his head bowed, eyes on the floor, and his shoulders hunched, like he was carrying all the sadness of the world.” His own sadness corresponded to his wife’s mood swings.

  In 1972, Rose Mary’s illness erupted cataclysmically. When she was flying high, she thought it would never end. Each high got higher, and the lows were the depths of hell. This time there was talk of a new experimental treatment, lithium. Paul, grieving, helpless, decided to study the biological effects of lithium using NMR methods. He learned all he could about manic-depressive disorders and their treatments. He learned that in some cases lithium was highly effective, but the mechanism was unknown. NMR, he decided, could be a means of finding out, so he submitted a grant application for this purpose and wrote a small paper.

  They decided to try the new experimental treatment. Rose Mary started on a high dose of lithium and was at first too groggy to function. Gradually, the dose was lowered, and she found herself symptom-free. From that day to this, she has never had another episode. “Lithium saved my life,” she says simply.

  But it didn’t save their marriage. Paul, she says, had put up with a lot, but by now whatever feelings he had had for her were gone. He was rarely home, and when he was home he was distant. She, too, was unhappy with the marriage. Paul hoped to begin entertaining at their home, and expected her to be able to do this. He had many foreign visitors and wanted them to stay at their house. She didn’t like these responsibilities and intrusions. This was one more mismatch driving them apart. Rose Mary wanted to repair the marriage; Paul saw no point in trying. Whenever something went wrong, she wanted to have discussions of the grand problems of their lives. He wanted only to deal with the matter at hand. They had arguments hot and cold. The cold, when Paul and Rose Mary would not speak to each other for days, burned deepest. Rose Mary feared no other woman because she knew he “was married to MRI.” She kept coming up with ideas about how to start over. Life continued in this calm but unsatisfactory manner for years. Then they separated for several years and finally, in 1984, divorced.

  If he was not exactly “married to MRI,” Paul certainly was preoccupied night and day with this new baby of his.

  7

  The Worldwide Laboratory

  Discovery consists of seeing what everybody has seen and thinking what nobody has thought.

  —Albert von Szent-Györgyi

  There never was an old theory of MRI; it was born whole. Paul once gave a talk titled “Magnetic Resonance Imaging: Why It Takes So Long to Understand Simple Things.” From 1946, once the phenomenon of NMR was understood any knowledgeable person could have invented MRI, but until Paul no one had the clarity of mind or the creativity to realize its beauty. Even so, everything new has precedents. All of the efforts to use projection reconstruction techniques for various purposes were mathematical precedents to Paul’s image reconstruction methods, although he was unaware of them and wound up reinventing them. A different case is that of the late Herman Carr, a professor of physics at Rutgers University, who in 1951 made what in current MRI would be called a “phantom,” three pieces of rubber of volumes 3:2:1, which were placed in a one-dimensional magnetic field gradient to simulate the chemical shifts of ethanol. The signals obtained from this phantom agreed with those from liquid ethanol itself, providing what may be the first one-dimensional NMR image ever recorded. This was twenty years before Paul had the idea of deliberately applying controlled field gradients to obtain multidimensional spatial images. Herman Carr wrote letters and articles to Physics Today and other publications complaining that his early work was overlooked.

  Dr. Carr was not alone in using field gradients for specific purposes. Unknown to Paul at the time, and perhaps to Dr. Carr, several people were using field gradients to identify separate physical positions of chemicals, and in unpublished work at the National Institutes of Health, Vsevolod Kudravcev (“Kud”) recognized that field gradients could be used to localize a volume of interest. All of these applications depended on the spatial variation of the NMR signal. Paul explained, “All of these were important precedents, but each was very specific to the [particular experiment] and with no intimation of generality in the process.”

  The important difference was that unlike earlier scientists, Paul was from the very beginning thinking of a general process. Although the citations often refer to two-dimensional imaging, that was a more practical way than the natural three-dimensional procedure, which was not as easy to implement both technically and mathematically. Two-dimensional imaging, Paul said, “was really a stopgap solution.” Or, as David Hoult succinctly declared, “You need three dimensions to win a Nobel Prize.”

  Other things are erroneously believed to be precedents to MRI. Advances in computer technology have made MRI clinically practical, but the concept of MRI did not involve computers. The first images were actually drawn by hand. It was the other way around: the advent of MRI, along with other imaging methods such as positron emission tomography and computed tomography (CT), led to the belief for the first time that computers might have an important role to play in medical imaging. Superconducting magnets, now used in clinical imaging, are these days
usually believed to be precedents and requirements for MRI. Not true: all of the early development of imaging was done using iron core magnets or ordinary electromagnets. In the early days there was a widespread belief that the high fields generated by superconducting magnets would actually degrade images.

  And then there is Raymond Damadian, at that time of the SUNY Downstate Medical Center in Brooklyn, who became a phenomenon not often seen in science. Dr. Damadian campaigned heavily for his priority in MRI through the press, and even forced a congressional hearing about supposed bias against him at the National Institutes of Health. As a result of this behavior he managed to share honors with Paul on three occasions, including induction into the Long Island Technology Hall of Fame and the National Inventors Hall of Fame, and receiving the National Medal of Technology. The issue really should have been put to rest by now since only two Nobel Prizes were awarded for MRI. There could have been three, and Damadian was not included.

  Dr. Damadian made an important contribution, which was to show that tissue water relaxation time, an important source of contrast in MRI, is longer in cancerous than in normal tissues. From very early days, even in the laboratories of Felix Bloch and Edward Purcell, people looked at proton resonances from their fingers and thumbs, body parts that could be inserted into standard machines. They also speculated about measuring the composition of tissues noninvasively using such techniques. The general opinion at the beginning of the 1970s, just before MRI exploded, was that relaxation times in excised tissues showed such wide variability that relaxation measurements could not be medically useful. Dr. Damadian’s finding in 1971 that relaxation times differ between malignant and normal tissues shifted the attention of the scientific community once more toward the question of what physiological or pathological properties water relaxation times could reveal. Was the phenomenon Damadian discovered real? The answer is “yes, but,” “sort of,” and “sometimes.” It is not a unique signature for cancer. Other diseases, and even normal activities such as exercise, also produce differences in tissue relaxation times. And we have discovered that these NMR parameters change whenever blood flow or swelling or dehydration occurs.