Portal de Periódicos da CAPES

Herman Yaggi Carr, PhD (1924–2008): A tribute

2009; Wiley; Volume: 29; Issue: 6 Linguagem: Inglês

10.1002/jmri.21772

1522-2586

Vivian S. Lee,

Advanced MRI Techniques and Applications

Herman Yaggi Carr was born on November 28, 1924, in Alliance, Ohio, to Robert Herman and Jeanette Yaggi Carr (Fig. 1). He served in WWII as an Army Sergeant in the 12th Weather Squadron Air Corps, which gave him some experience with a pulse apparatus that later proved quite useful. He was married to Hilda Hagen Kinney, who passed away in 1986, and is survived by 2 children and 3 grandchildren. His daughter, Amanda Carr Sozer, a distinguished forensic scientist, spoke at the Toronto ISMRM meeting, sharing reflections on her father's life as a father, scientist, and humanitarian. Herman Carr. Carr attended Harvard College as an undergraduate, where he graduated summa cum laude in 1948. He continued as a graduate student in physics, supervised by Edward Purcell. He received his PhD in physics from Harvard in 1953. After graduation, he spent his career at Rutgers University, where his research moved in new directions, focusing on xenon and the study of fluids and the liquid–vapor critical point. Together with Richard Weidner, he authored a physics textbook entitled, “Physics from the Ground Up” (1). In 1987, he retired from Rutgers. On April 9, 2008, less than a month before he was scheduled to deliver the 2008 Lauterbur Lecture at the ISMRM, Herman Carr passed away in his home in Bridgewater, New Jersey. Serendipitously, approximately 4 years ago, I had the opportunity to meet Herman Carr. He contacted me because he was seeking to understand more about where the field of magnetic resonance had moved over the past 50 years, and he was curious about the clinical impact it was making. This led to a short but memorable relationship which included a special grand rounds presentation by Carr at NYU Langone Medical Center on May 29, 2006, entitled, “Magnetic Resonance from the Gradient Up,” an allusion to the title of his earlier physics textbook (1). The material in his presentation, together with some other archival materials that his daughter shared with me and a historical piece written by Carr in 1996 (2), form the basis of this tribute to Herman Carr for the many contributions he made to our field during the 1950s when his work focused on nuclear magnetic resonance research. As Carr tells it, in early November 1949 at the beginning of his second year as a graduate student at Harvard, the bulletin for the American Physical Society meeting (3) arrived with the abstract of Erwin Hahn's contributed paper that first described the phenomenon of the “spin echoes” (4). Purcell suggested that Carr read the abstract and try to understand this fascinating effect. Over the Christmas holidays, Carr was in Urbana, Illinois for a conference and made a visit to Hahn's laboratory. With this trip, Carr discovered a passion for the spin echo that led to an incredibly exciting and productive 3-year period of graduate studies with Purcell. In a letter dated November 14, 1950, Erwin Hahn starts out, “Dear Carr, Well! Well! It gratifies me to hear you now have echoes at Harvard.” And he closes with “I am happy to have you be so interested in spin echoes” (5). Indeed, the contributions Carr and Purcell made in the early 1950s truly amplified and transformed the understanding of spin echoes across a range of explorations. Most of the results of Carr's graduate work were published with Ed Purcell in a landmark article in Physical Review in 1954, entitled, “Effects of Diffusion on Free Precession in Nuclear Magnetic Resonance Experiments” (6). As he tells the story, it was during lunch one day, while explaining Hahn's spin echoes to another graduate student, that Carr realized that the explanation could be further simplified by considering two unequal pulses—a 90° pulse followed by a 180° pulse (Fig. 2). Moreover, it was easy to understand how a sequence of echoes could be induced by following one initial 90° pulse with a series of 180° pulses. When Carr told Purcell about this idea later that afternoon, he recalls Purcell's delighted response “Elegant!” (2). The formation of a 90°–180° spin echo. Reprinted figure with permission from Carr and Purcell. Phys Rev 1954;94:630–638. Copyright 1954 by the American Physical Society. The experiment could be repeated several times, varying the time between 90 and 180° pulses (Method A), or, it could be conducted following a single 90° pulse and a train of 180° pulses (Method B) (Fig. 3). For viscous samples with negligible diffusion, Carr and Purcell knew Methods A and B (Fig. 3) should give similar results. But using a water sample, Carr showed that the decay is dominated by an exponential, exp(−kt3), with a time constant of only 0.2 s for Method A; while using a train of 180° pulses (Method B), the decay has a simple exponential decay with time constant of 2.0 s (Fig. 4). While in general any two-pulse echo decays because of T2 and diffusion, with this elegant experiment, Carr and Purcell demonstrated the effectiveness of the 180° pulse train in partially eliminating the decay caused by diffusion, so that “The envelope of these echoes indicates the decay of polarization in the equatorial plane,” or T2 relaxation. Comparison of two methods for observing transverse decay. Reprinted figure with permission from Carr and Purcell. Phys Rev 1954;94:630–638. Copyright 1954 by the American Physical Society. Method A (see Fig. 3) associated with water at 25°C shows decay with a time constant of approximately 0.2 s, where decay is largely determined by molecular diffusion and dominated by the factor exp (−kt3). Method B (see Fig. 3) associated with water at 25°C shows decay with a time constant off approximately 2 s, the effects of diffusion minimized with the train of 180° pulses. Reprinted figure with permission from Carr and Purcell. Phys Rev 1954;94:630–638. Copyright 1954 by the American Physical Society. Interestingly, as Carr describes, the use of gradients in NMR was in its infancy—note the early description in this article (6). “A gradient G may be obtained at the sample by placing symmetrically on either side of the sample two long current carrying wires or two circular turns of wire. The current directions should be such that the fields oppose.” And also note the allusion to field inhomogeneity problems: “… it is necessary that this G be large compared to the average gradient of the field due to the magnet. To satisfy this condition, it is wise to search for a very homogeneous spot in the magnet.” Figure 5 shows one of the earliest Maxwell (or anti-Helmholtz) gradient pairs that Carr used, when at Rutgers, which has been contributed to the ISMRM historical archives collection. One of the earliest Maxwell (or anti-Helmholtz) gradient pairs that Carr used when at Rutgers. Also, the 1954 study (6) gives consideration to the effects of convection or flow and an observation and explanation of what we call even-echo rephasing, or as he also referred to it, “odd-echo dephasing.” Figure 6 shows that the odd-numbered echoes are much smaller than the even-numbered ones. Carr goes on to conclude that the observation of the larger even-echo amplitudes can be used as an indicator of the presence of convection. The effect of molecular convection on Method B (see Fig. 3) decay, illustrating what is commonly referred to as “even-echo rephasing.” Reprinted figure with permission from Carr and Purcell. Phys Rev 1954;94:630–638. Copyright 1954 by the American Physical Society. Lastly, in the final section of this same paper (6), Carr and Purcell described what happened if you preceded the 90° pulse with a 180° pulse, as we now use in inversion recovery imaging, and showed: “that T1 may be calculated directly from the measured value of τnull by using the relation τnull = T1 ln2.” The impact of this 1954 paper was tremendous. By 1961, the paper had been cited even more times than Hahn's original spin echo paper. In 1977, Carr received a letter from the Science Citation Index, which informed him that the 1954 Physical Review paper was among the 500 most cited during the years 1961–1974 (7). In the 1950s, the electronic techniques available for Fourier transforms had not been displayed. Nevertheless, the Fourier transform relating the time dependence of the FID following a pulse and the frequency distribution of precessing spins was recognized early, without computers (8). Carr noted “Very early we used our own brains to make speedy Fourier transforms of at least the simple FIDs we were observing” (2). Even when three spin groups were involved, experienced physicists like Carr could recognize the FID pattern reasonably well, provided the lines had equal intensity. This mental agility served him well in this set of experiments that he published with Purcell in 1952 (9). This work was preceded by a publication earlier in 1952, where Purcell and Norman Ramsey had successfully explained the mechanism for J-coupling as an indirect electron-coupled nuclear spin–spin interaction. Carr and Purcell, using hydrogen-deuterium (HD) gas, showed that the modulation pattern in the “tail” in the D resonance in HD had a simple beat pattern of two signals of equal amplitude, reflecting the two spin groups that corresponded to the two possible orientations of the neighboring proton magnetic moment (Fig. 7a). For the protons, they saw and recognized the typical beat pattern for three signals of equal frequency separation and amplitude—the three groups corresponding to the three possible orientations of the neighboring deuteron magnetic moment (Fig. 7b). Benefitted by their agility with “mental” Fourier transforms, Carr and Purcell were able to experimentally verify the basis of J coupling. Experimental verification of the basis of J-coupling. A: Deuteron signal from HD. B: Proton signal from HD. Reprinted figure with permission from Carr and Purcell. Phys Rev 1952;88:415–416. Copyright 1952 by the American Physical Society. Some “mental” Fourier transforms were not too difficult, but as Carr tells it, most visitors to Carr and Purcell's lab had a lot more trouble understanding the FIDs from the then much-discussed ethyl alcohol. It was recently discovered that a chemical shift involved three proton lines with intensities in the ratio 3:2:1, reflecting the methyl, ethyl, and hydroxyl protons, respectively. From their diffusion studies, using the gradient coils as shown in Figure 5, Carr had made the connection that resonant frequencies could be modified based on location in the gradient. That is, he observed each position in the gradient corresponded to a different resonance frequency. Therefore, to help visitors understand the ethyl alcohol FID, he constructed a phantom that he called “synthetic alcohol.” It consisted of three small pieces of rubber with strong proton resonance placed in the coil at positions such that the separation in space was in the same proportion as the ratios of differences of the resonance frequencies for the three lines of the ethyl alcohol resonance. Moreover, the three small pieces of rubber were cut in pieces, which had lengths in the ratios of 3:2:1 to correspond to the alcohol line intensities. After applying a 90° radiofrequency pulse, the FID in the time domain looked very much like the real alcohol FID (Fig. 8) (2, 10). Comparison of FIDs from ethyl alcohol (real, upper left) and ethyl alcohol phantom (synthetic, lower left). Fourier transform of the real ethyl alcohol FID is shown in the upper right. For the phantom, H1, H2, and H3 are signals from small rubber tubes cut in the volume ratio 1:2:3 (lower left). Modified from Carr HY. “Free Precession Techniques in Nuclear Magnetic Resonance” PhD Thesis, Harvard University, Cambridge, Massachusetts, 1952, with permission from Amanda Carr Sozer. [Color figure can be viewed in the online issue, which is available at http://www.interscience.wiley.com.] With this simple experiment, published only in his Harvard PhD thesis (10), Carr used a gradient and spatial location within the field to simulate a frequency distribution. Carr's contributions to MR did not stop when he moved to Rutgers as a young faculty member in 1952. His work in driven equilibrium (11) and steady-state free precession (12) made significant fundamental observations that are highly relevant in the MR imaging community today. At Rutgers, Carr focused his work on the liquid–vapor critical region. He and others showed that, even in the absence of gravitational complications, the shape of the liquid–vapor coexistence curve could not be described by a pure power law with a single critical exponential (13). His group developed a temperature quenching technique to enable work even closer to the critical point, despite gravitationally induced density gradients. In 1987, Carr retired from Rutgers, although he continued to maintain a laboratory there until his death. As the invited 2008 ISMRM Lauterbur lecturer, Herman Carr was preparing to deliver a presentation entitled “The Origin of MRI in Physics: Curiosity, Serendipity, and Inventions.” The Society would have been graced by his presence and would have gained memorable insights into the earliest days of NMR that would have been enlightening and provocative. At the same time, our hope was that for Carr, the opportunity to be exposed to the breadth and depth of applications that his original contributions have spawned would have made the meeting especially meaningful for him. From the laboratory to the MRI clinic, introduction of concepts of the 90°–180°-spin echo, the Carr-Purcell sequence of a train of echoes, improvements in T1 and T2 quantification, inversion recovery, diffusion-weighted imaging, steady-state free precession, among others, are fundamental concepts of NMR and now MRI to which Carr made key contributions and earn him the recognition as one of the founders of magnetic resonance in medicine. Every time an MRI is performed, his memory lives on. In the words of Nikola Tesla (14), “The scientific man does not aim at an immediate result. He does not expect that his advanced ideas will be readily taken up. His work is like that of the planter—for the future. His duty is to lay the foundation for those who are to come, and point the way. He lives and labors and hopes.” Herman Carr was an extraordinary scientist and a committed humanist. The impact of his work on the field of magnetic resonance in medicine has been immeasurable. I thank Professor Erwin Hahn for his thoughtful recollections and insightful improvements to the text.