| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorials |
1 From the MR Laboratory, Mayo Clinic, 200 First Street SW, Rochester, MN 55905. Received January 19, 2004; accepted January 21. Address correspondence to the author (e-mail: riederer@mayo.edu).
Index terms: Editorials • Magnetic resonance (MR) • Radiology and radiologists • Radiology and radiologists, history
On October 6, 2003, the Nobel Foundation announced that the two recipients of the Nobel Prize in the category of physiology or medicine were Paul Lauterbur, PhD, of the University of Illinois, and Sir Peter Mansfield, PhD, of the University of Nottingham, England. Drs Lauterbur and Mansfield, along with the awardees in other categories, were honored at the Nobel Awards Ceremony in Stockholm, Sweden, on December 10, 2003. Much has already been written about this year’s Nobel Prize, the accomplishments of the awardees, and indeed the accomplishments of numerous scientists who have advanced the field of magnetic resonance (MR) imaging. The 2003 Nobel Prize not only is a recognition of the pivotal achievements of Drs Lauterbur and Mansfield, but it also can be interpreted as a recognition of the overall benefits provided to society by MR imaging and high-technology medical imaging in general.
Dr Paul Lauterbur’s principal contribution was his demonstration of the capability to use nuclear MR, which was discovered in the middle 1940s, to form an image. The ingenious technical advance was his recognition and subsequent demonstration of the fact that by adding a magnetic field gradient to an underlying large static magnetic field, one could make the resonant frequency of an object spatially dependent, and, thus, an image of that object could be formed (1). Thus, the strength of a nuclear MR signal at a point in space could be determined by measuring the signal intensity at the corresponding frequency. This seemingly counterintuitive association between frequency and spatial position is now accepted as second nature to those who use MR imaging and is the basis of virtually all methods of MR image formation.
Since the time of Dr Lauterbur’s contribution 30 years ago, MR imaging has advanced from a method that was used in a handful of specialized physics laboratories worldwide to an everyday, widely accepted examination that is used for clinical and research applications. Although the installation of large superconducting magnets in community hospitals might have once seemed a pipe dream, today it is a reality. MR imaging data from 2002 indicate that there are almost 15,000 whole-body MR imaging units installed worldwide and that almost 7,000 of these units are in the United States (2).
MR imaging is the primary imaging modality used to examine patients who have signs, symptoms, or concerns regarding neurologic disease. Many neuroradiologic imaging examinations that used to be commonly performed, such as conventional angiography and myelography, have been largely supplanted by counterpart MR imaging examinations. MR imaging has also become the "go-to" imaging modality for many concerns regarding the musculoskeletal system and increasingly for concerns regarding the abdomen, pelvis, and cardiovascular system. There is no doubt that MR imaging has had a major influence on modern medicine, and much of the effort in making this so has been chronicled in the pages of this journal.
Dr Peter Mansfield was one of the leaders of the MR imaging research effort at the University of Nottingham in the 1970s. His contributions are numerous, dating back to studies of the structure of crystals with nuclear MR (3). He is well known and recognized for his development of echo-planar MR imaging (4). The ingenuity of this method was the capability to acquire all of the data necessary to form an image in a very short time—on the order of 50 msec—by rapidly oscillating the magnetic gradients to sample so-called k space. Although the theoretic basis for echo-planar imaging was sound, in some sense the idea was years ahead of its time. If one were to have considered in the early 1980s the possible acquisition times required for the then-described MR imaging techniques, echo-planar imaging would have been at the extreme fast end of the scale, with acquisition times (in tens of milliseconds) several magnitudes faster than those required for the conventional T1- and T2-weighted spin-echo techniques (in minutes and tens of minutes, respectively).
Although the short acquisition times were attractive conceptually, the use of echo-planar imaging in the middle 1980s led to a number of severe image-degrading effects, including poor signal-to-noise ratio (SNR), poor spatial resolution, and general blurring, particularly of objects containing fat. Also, the technical performance capability of the magnetic gradients that were available at that time was inadequate in terms of possible amplitude and switching times.
Two general scientific research trends emerged from this scenario: One trend was to look for ways to fill the gap in acquisition times between milliseconds and minutes. Specifically, the goal was to maintain the quality and contrast behavior of standard spin-echo MR images while targeting acquisition times within the approximate 1-second to 1-minute range. Several important advances in MR image acquisition emerged as a consequence of this goal. One was the development of gradient-echo MR imaging (5), which allowed repetition times in the 10–100-msec range and thus image acquisition times as short as 1 second. Today, gradient-echo sequences are used predominantly for T1-weighted MR imaging and because of the short repetition time required are the method of choice for many kinds of three-dimensional acquisitions, such as those in MR angiography (6–8). A second important result of this search for not-quite-as-fast alternatives to echo-planar imaging was the introduction of rapid acquisition with relaxation enhancement, or RARE, MR imaging (9), which is also sometimes referred to as fast spin-echo or turbo spin-echo MR imaging. The original version of RARE MR imaging and its multirepetition variants (10) have become workhorses of clinical MR imaging. Both gradient-echo and RARE spin-echo MR imaging methods are used in literally tens of thousands of clinical examinations daily and are here to stay.
The second general trend was the undertaking of efforts to address many of the factors that limited the implementation of echo-planar MR imaging and to some extent limited virtually all MR image acquisition methods. One example of these efforts was the development of high-quality receiver coils for improved SNR. The first such development was intended to improve the homogeneity of the response with use of a single composite coil (11). Several years later, this was further advanced by the use of multiple coils and receiver channels (12) to provide improved SNR. Another major technical improvement was the development of shielded gradient coils (13), which ensured that the k-space region being sampled during a measurement was actually the region that was intended. Yet another major engineering improvement was that of the receiver chain, which was upgraded by MR imaging manufacturers to accommodate very high data acquisition rates, typically 10 times faster than the rates achievable with the then-standard methods.
Perhaps the most important technologic advance in MR imaging has been the gradual improvement in magnetic gradient systems themselves: Maximum allowable amplitudes have increased from 10 to 40 mT/m or more, and rise times (which generally represent wasted time) have decreased from 0.5 to 0.1 msec or less per k-space line. These advances have facilitated the acquisition and processing of MR imaging data more quickly, more accurately, and with improved SNR and thus to precisely address the limitations of echo-planar MR imaging.
In addition to the technology enabling the acquisition of high-quality echo-planar images that is becoming available, an additional driver has been the identification of MR imaging applications that require very-high-speed acquisition. Perhaps the prime example is diffusion-weighted MR imaging, a method first presented in 1986 (14). With this technique, random motion of the nuclear spins is determined by first dephasing the spins by using a high-amplitude, long-duration gradient and then rephasing the spins with a second gradient that has an identical but reversed amplitude-time product. The diffusion signal is based on the failure of all spins to rephase exactly, a consequence of their random motion or diffusion. This causes signal attenuation. The problem with this technique is that acquisitions performed by using long-duration, high-amplitude gradient pulses are particularly prone to additional artifacts from bulk non–diffusion-related motion, and this limitation plagued early attempts to obtain diffusion-weighted MR images. The ideal solution was to incorporate the diffusion-encoding gradients into a very rapid MR image acquisition examination—that is, echo-planar imaging (15).
Echo-planar imaging has made diffusion-weighted imaging a clinical reality. Today, a substantial percentage of clinical MR imaging examinations include a diffusion-weighted sequence. For example, in our neurologic MR imaging practice at Mayo Clinic, approximately 20% of brain MR imaging examinations include a diffusion sequence. For diffusion encoding along all three axes, a multisection echo-planar sequence can be performed in 40 seconds. More recently, investigators have been turning to measurement of the diffusion tensor (16), in which the orientation and magnitude of diffusion are estimated at each point to determine, for example, white matter tracts (17). It is fair to say that echo-planar imaging has been a technical enabler of diffusion MR imaging.
In addition to its influence in medical imaging, MR imaging has had a profound effect in other fields. Since functional neurologic MR imaging techniques were first presented in the early 1990s (18–21), there has been a virtual explosion in the study and application of these techniques. With use of this technique, specific loci of brain activity are probed and identified by exposing the patient to some kind of stimulus, such as visual or sensory stimulation. Brain activity is localized indirectly by detecting subtle but measurable differences in blood flow. Additionally, MR signal intensity is modulated by changes in the relative amounts of deoxyhemoglobin versus oxyhemoglobin (22), with the decrease in deoxygenated blood leading to a signal intensity increase. Although functional MR imaging can be used clinically—for example, to aid in the planning of certain neurosurgical procedures—this modality has extended well beyond the field of diagnostic imaging.
The ability to use MR imaging to noninvasively probe the individual regions of the brain that control vision, sensation, motor function, memory, language, and other processes has made this an extremely valuable modality to those engaged in virtually any kind of brain-related research. To get an idea of the scope of functional MR imaging applications, a recent search in the National Library of Medicine’s PubMed database of biomedical journals resulted in the identification of over 40,000 publications related to the field of functional MR imaging of the brain.
Another area of increased use of MR imaging is animal studies. In particular, MR imaging can be used in a noninvasive way and thus represents a procedural shift from the manner in which many animal studies have been typically performed in the past—that is, with animal sacrifice required to monitor and verify each phase of a study. With MR imaging, the progress of an induced disease or condition (possibly genetically controlled) or the effect over time of an administered drug can now be studied by examining substantially fewer animals longitudinally; thus, sacrifice is no longer required in many cases. Investigators have adapted MR imaging systems to the special demands of animal imaging with smaller bore magnets, higher capacity gradients, finer spatial resolution acquisitions, and monitoring tailored to the physiologic features of animals (23). This is not to say that animal imaging is the exclusive domain of MR imaging. Small-bore computed tomography (CT) and micro–positron emission tomography also are being used for this purpose (24), with the choice of modality selected according to the unique physical parameters that each technique provides.
MR imaging and CT are often used as examples of how modern technology is having an influence in medicine. The increased clinical use of MR imaging and CT is illustrated in Figure 1; the numbers of patient examinations performed with these two modalities during the past 3 decades at Mayo Clinic are shown. During the past 10 years—that is, from 1992 to 2002—the annual growth rate of MR imaging use has been 10.4% and that of CT use has been 8.3%. The purpose of presenting these results is not to suggest the superiority of one modality over the other but rather to show that both modalities have been proven to be important clinically and that their use continues to grow rapidly. The increased use of MR imaging mimics that of CT; in fact, if one displaces the MR imaging results 11 years to the left, they almost exactly match those of CT.
|
The increased use of "high-tech" diagnostic imaging in general is illustrated in Figure 2. The presented data are the percentages of radiologic examinations performed at Mayo Clinic by using high-tech (defined here as CT, ultrasonography [US], and MR imaging) versus non–high-tech (defined as all other imaging modalities) imaging modalities, dating back to 1970, before CT was introduced. This author recognizes that these definitions are arbitrary, as technology has clearly had an influence in virtually all other imaging modalities, a prime example being digital subtraction angiography, as reported by Kruger and colleagues (25) more than 2 decades ago. Turning back to Figure 2, from the presented data it is seen that as of 2002, almost 30% of all imaging examinations were performed by using these high-tech modalities. Moreover, this percentage is growing. In fact, extrapolation of the data indicates that most radiology departments will be totally dominated by CT, US, and MR imaging procedures shortly after the year 2030.
|
|
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
