Looking Within: How X-Ray, CT, MRI, Ultrasound, and Other Medical Images Are Created, and How They Help Physicians Save Lives / Edition 1

by Anthony Brinton Wolbarst

Overview

A hundred years ago, a doctor had no way to look within the body of a patient other than to slice it open. That changed radically at the turn of the century, with the discovery of X-rays. X-ray and other forms of diagnostic imaging technology developed slowly but steadily from then until the 1970s, at which point a revolution occurred. Made possible largely by the availability of powerful but inexpensive computers, the rapid and widespread adoption of computed tomography (CT) and, a decade later, of magnetic resonance imaging (MRI) greatly expanded the power of clinical imaging, and even changed the ways in which physicians view and think about the human body.

This unique guide explains how the principal imaging devices work and how they help physicians save lives. It gives readers a grasp of the major medical technologies that might come to play important roles in their lives, and it provides succinct, easy-to-understand, and reliable explanations for those who wish to explore the issues of the associated benefits, costs, and risks in an informed manner. In nonspecialized language, Looking Within discusses how X-ray, fluoroscopic, CT, MRI, positron emission tomography (PET), ultrasound, and other medical pictures are created, and explores the essential roles they play in the diagnosis and treatment of patients. It should be of interest to patients and their friends and loved ones, and to those who are simply curious about this vitally important, exciting, and cutting-edge branch of medicine. Its brief but clear descriptions of how these essential tools work should also be of value to health care providers in supporting and educating their patients.

Product Details

ISBN-13:	9780520211827
Publisher:	University of California Press
Publication date:	11/16/1999
Edition description:	First Edition
Pages:	219
Product dimensions:	7.00(w) x 10.00(h) x 0.60(d)

About the Author

Anthony Brinton Wolbarst, a physicist formerly at Harvard Medical School and the National Cancer Institute, is currently at the U.S. Environmental Protection Agency and an Adjunct Associate Professor at Georgetown University Medical School. Dr. Wolbarst is the author of Symmetry and Quantum Systems: An Introduction to Group Representations (1977) and Physics of Radiology (1993), and editor of Environment in Peril (1991).

Read an Excerpt

Chapter One: From the Watching of Shadows

"Our science is from the watching of shadows." Ezra Pound, Canto 85

One hundred years ago, the German physicist Wilhelm Conrad Roentgen (figure 1) happened upon X rays. Although no one realized it at the time, this most extraordinary and mysterious discovery foreshadowed the quantum upheavals that would turn the physical sciences upside down in the early decades of the twentieth century. More immediately and spectacularly, though, it flung open a door that led into a new and completely unanticipated dimension in the practice of medicine-the ability to look within a patient's body without having to slice it open.

"A New Kind of Ray"

It is difficult, from today's vantage point, to imagine just how primitive medicine actually was a century or so ago. The idea of surgery can still make anyone a little nervous, but in earlier days, it elicited feelings of sheer horror. Many are the stories of the wounded or seriously ill who begged for a hard blow to the head, or even death itself, rather than the ordeal of the knife. Alcohol and opium could induce some degree of numbness, but anesthetics-the kind that really knock you out, like ether and chloroform-were not in wide use before the American Civil War. The painkilling properties of nitrous oxide gas and ether had been known since around 1800, but deep anesthesia was not used in surgery until midcentury.

In those days, moreover, something like half of all patients who did endure the amputation of an arm or leg died of infection anyway. It wasn't until the 1860s that Louis Pasteur, Joseph Lister, Florence Nightingale, and others recognized and began preaching the importance of cleanliness, disinfectants, and sterilized surgical instruments. But even with these improvements in the operating room (which, like anesthesia, took decades to gain general acceptance), death from infection was commonplace until penicillin and other antibiotics became readily available after World War II.

Medical diagnosis, too, was almost entirely art, and little science. The physician could measure body temperature, blood pressure, pulse rate, and a few simple chemical attributes of blood and urine, but not much else. Odors and subtle aspects of a patient's appearance commonly provided equally important clues. But a doctor often had no way to know what was going on within the body other than to cut it open.

All of that changed overnight in 1895. Roentgen, a respectable but rather obscure professor at the University of Würzburg, had been experimenting with an apparatus of widespread scientific interest at the time, a vacuum tube through which electric charges were flowing. Late in the evening of November 8, working in a darkened room, something unusual caught his eye: when an electric discharge occurred in his tube, a nearby piece of paper coated with a chemical compound of barium, platinum, and cyanide produced a glow. With his glass tube completely enveloped in black cardboard, no light from it could be reaching the coated paper. So something invisible had to be passing through the cardboard and reaching the barium platinum cyanide, inducing it to give off light. Roentgen had, in fact, discovered X-ray radiation by observing X-ray fluorescence (the emission of light caused by an X-ray stimulus) in a nearby fluorescent material.

By placing various objects between the tube and his fluorescent screen, Roentgen learned that they affected the brightness of the emitted light by different amounts. Paper and cardboard had little effect, but a thick sheet of metal quenched the light completely. And when he held his hand in the path of the X-ray beam, he could make out the bones of his fingers projected in silhouette upon the screen. A short while later, Roentgen produced the first X-ray record, capturing for all time his wife's hand and signet ring on a glass photographic plate (figure 2).

Word of this wonder spread like wildfire, and the experiment was easy to reproduce. Within months, physicians throughout the world were using X-ray images to extract shrapnel and set broken bones. Roentgen had discovered "a new kind of ray, " as he described them, and in so doing he had created a splendid window for looking within the living body.

For the better part of a century thereafter, innovations in the field of medical imaging came slowly but steadily, and a few were quite remarkable. But the advances in recent years have been as revolutionary as the computers that have made them possible. If the people who developed the automobile, the airplane, the telephone, or the television were to run across modern versions of their inventions, they would probably understand a great deal of what they found. But if Roentgen were to wander through a medical imaging center today, with its computed tomography (CT) scans, magnetic resonance imaging (MRI), and positron emission tomography (PET), much of what he saw would mystify him.

Looking Within will walk Herr Professor Roentgen, and anyone else who would like to come along, through a modern imaging department, explaining the medical marvels that would so amaze him. But let's start with a technology Roentgen should feel quite comfortable with, and show how an ordinary X-ray film study of a hand is produced and used today.

X-Ray Film of a Cracked Bone

When a patient shows up at her door, a physician will listen to the symptoms, do a physical examination, and perhaps take some blood or urine specimens (which today can provide highly specific and valuable information). From her interpretation of the results, she can probably limit the diagnosis to one or a few possibilities. Medical imaging may now step in to play a decisive role in confirming, refining, modifying, or refuting the initial diagnosis. Imaging may also be invaluable in planning the treatment and in following the patient's progress over time.

It may not be readily apparent to the patient, but this general process of gathering relevant information on the malady, considering the possible explanations, and focusing first on the most likely ones goes on quietly even with as simple a problem as a broken bone.

Kathleen Nealon, the sixteen-year-old star pitcher for her high school softball team, took a hard blow to the left hand from a batted ball, causing a great deal of pain and rapid swelling. Her older sister, Kelly, who had dropped by to watch part of the game, drove her to the emergency room of the local hospital.

After carefully inspecting the hand, the emergency room physician sent Kathleen to the radiology department for an X-ray film. An imaging study was needed for a correct diagnosis that would, in turn, guide Kathleen's treatment. If no bones were damaged, Kathleen could get by with elevation of her hand, intermittent application of a cold pack, and medication to reduce swelling and discomfort. If the radiologist found a hairline crack, the hand might need a cast to counteract any stresses on the injury during healing. If a bone had been broken into separate pieces, it might even be necessary to wire them together surgically for proper setting. Before there were X-ray films, a physician would have had to stabilize a bone without being able to see clearly how to position the pieces, and that could result in weakness and deformity after mending.

Kathleen's X ray took less than five minutes. The radiographer (also known as a radiologic technologist) positioned her swollen hand on a cassette, which contained a sheet of radiographic film, adjusted the height of the X-ray tube above it, and reduced the dimensions of the rectangular X-ray beam until it barely covered the hand (figure 3a). Then he protected Kathleen's body and neck with a lead-lined apron, which strongly absorbs any stray X rays. He stepped behind a shielding wall, set the controls of the X-ray machine, and kept watch on his patient through a lead-glass window as he shot the film. He then replaced the exposed film in the cassette with a fresh one, repositioned the hand, and took a second film.

In a few minutes, the films were developed and ready for inspection (figure 3b). Both were of high enough quality for the radiologist to identify the problem and guide Kathleen's treatment. As with most radiographic studies, the contrast between bone and soft tissue was high. There was almost no visual noise interfering with what had to be seen. And the films had sufficient sharpness and resolution of detail to reveal a clean, simple break in one of the bones into two separate pieces. These had not been displaced relative to one another, but the radiologist recommended a cast anyway, to ensure that the bone would be rigidly immobilized.

A month later, when the bone was nearly healed, the cast came off. A few weeks after that, Kathleen was back in there pitching.

What a Physician Needs from a Medical Image

My wife and I live in Washington, and even though we visit the Lincoln Memorial with out-of-town friends two or three times a year, we never tire of it. We love it for the strength and integrity carved into the wonderfully lifelike face of the man, along with his gentle understanding and acceptance of human frailties (figure 4a). It's all right there, in the stone.

Another, and altogether different, visual treat is Marcel Duchamp's Nude Descending a Staircase (figure 4b). Duchamp interprets the human form by portraying its flow, and his painting is as abstract and impersonal a representation as Lincoln is solid and familiar.

Some would argue that Lincoln is a more important work, of greater inherent value because of its directness and traditional authenticity. No one ever has to ask what it means. Such a perspective would miss a crucial point: the two representations are intended to do very dissimilar things, and both succeed splendidly in their individual ways of depicting reality-but an informed interpretation is essential for a true appreciation of either.

The situation is much the same for a medical image. You might suppose that the value of a picture increases with its visual similarity to the part of the body that it examines. But much of that information content will invariably be medically irrelevant at best, or detract from or even hide the diagnostically critical features. Take a look at the studies by X ray, nuclear medicine, and magnetic resonance angiography (all of which we shall discuss later) in figure 5, and you'll see that what a physician has to work with can be quite distinct from photographic reality. What is important is what he can "read" in the image. His job is to detect any significant anomaly in it, and identify a corresponding irregularity in the patient's body. He must then interpret this in terms of a deviation from normal anatomy or physiology-the what, how, and why of what has actually gone wrong with the cells, tissues, and organs. Only then, after settling on at least a tentative diagnosis, is it possible to choose the best treatment. A medical image will be considered good if it helps make any or all of this happen reliably and easily. A diagnostic imaging system must therefore be able to display the specific, distinctive aspects of a patient's anatomy or physiology that are causing a problem, and be sensitive enough to pick up even very faint signs of it.

The specificity, sensitivity, and other characteristics of the various imaging tools, in turn, are determined by how they work-and they work in remarkably disparate ways. But although the several imaging technologies use quite different physical processes in carrying out their appointed tasks, as we shall see, they do share a fundamental commonality of approach: they create medical images by following and recording, by some means, the progress of suitable probes that are attempting to pass through a patient's body. The body must be partially, but only partially, transparent to the probes. If the probes all slip right through bones and organs without interacting with them, like light through a pane of clear glass, no differences among the tissues can be visualized. Similarly, if their passage is completely blocked, nothing much shows up. But if we choose probes that are only somewhat absorbed, scattered, reflected, delayed, or otherwise affected, we may be able to detect small differences in how they interact with different biological materials. And these differences can then serve as the raw material for the creation of diagnostically useful pictures.

When a uniform beam of X rays entered Frau Roentgen's or Kathleen Nealon's hand, for example, the bones and muscles attenuated it (i.e., removed energy from it, reducing its intensity) by different amounts, thereby casting a distinctive pattern of X-ray shadows in it. The no-longer-uniform beam that emerged from the hand then fell upon and exposed a photographic plate or film cassette. Finally, the X-ray shadow pattern was distilled into a permanent visual record when the photographic plate or film was developed.

Mammographic radiography, nuclear medicine, magnetic resonance imaging, and ultrasound use different physical probes in examining the body. These probes interact with the tissue immediately around them, and the nature of that interaction can be highly sensitive not only to the specific physical characteristics of the tissue, but also to the nature of the probe. It should be no surprise, then, that each imaging technology, with its own particular kind of probe, is suitable for the study of only certain kinds of medical problems. A fine crack in a small bone that would not show up at all with ultrasound imaging (which uses high-frequency sound waves) or with magnetic resonance imaging (magnetic fields and radio waves) may be fully visible in an ordinary X-ray film and perhaps in some kinds of nuclear medicine studies as well (gamma rays). Conversely, subtle differences among the various soft tissues of the abdomen that cannot be seen with the X rays of radiography or even CT may be easy to spot with ultrasound or MRI. Different probes, different interactions with the tissues, and different means of detecting the probes give rise to different images conveying different types of clinical information.

Through a Glass, but Not Too Darkly

Selecting the technology that employs the most appropriate probe is only the first step. Given that the physician chooses a suitable diagnostic test, the resulting pictures will be of little clinical use unless they are of good enough image quality. Although there are other critical factors as well, the three gold standards by which images (and the imaging systems that produce them) are most commonly judged are contrast, resolution, and visual noise.

When contrast is good, significant physical differences among the tissues show up as substantial differences in shades of gray (or color) in the image. The contrast between bone and the soft tissues of muscle or the internal organs, for example, is almost always strong in a radiograph, even in Roentgen's first plate-but there is little inherent radiographic contrast among the organs themselves, so it is sometimes diÐcult to see them at all. The same organs might show up with dazzling contrast, however, with an MRI or a nuclear medicine scan.

Some kinds of investigations, such as the search for calcifications in the breast (tiny flecks of bonelike material that may sometimes be suggestive of cancer), require high resolution (also called sharpness), the ability to display fine detail. X-ray films tend to provide extremely good resolution, and tiny objects and linelike features of interest within the body (such as the crack in Kathleen's finger, figure 3b) may show up well in a radiograph or mammogram. One important source of unsharpness is the blur introduced by patient movement-which is the reason for the inevitable "Take a deep breath and hold it!" that accompanies chest films. With some kinds of imaging, such as ultrasound and nuclear medicine, the power of resolution is inherently not great; but although those technologies may get low marks in visualizing anatomic detail, they do display the contrast needed to provide other (sometimes much more important) sorts of information on the patient's medical condition-such as how healthy the tissues of a particular organ are.

Visual noise refers to anything that interferes, a little or a lot, with an image, just as static noise from lightening in a storm will degrade a radio broadcast. An all-too-familiar example of visual noise (before the advent of cable) was the irritating snow that blew across your TV screen whenever the signal was too weak. But noise may assume more subtle forms, as anyone who has enjoyed an afternoon in the park with Georges Seurat well knows. His Sunday on La Grande Jatte is composed of countless little dabs of paint of various colors. From a distance, the image seems quite smooth and realistic (figure 6a), but the finer features are indistinguishable. At close range, the size of the individual dabs causes the picture to take on a speckled texture (figure 6b), and that, too, limits the amount of sharpness possible. Fine detail is not what Seurat had in mind, but the objectives of medical imaging are different. A digital image, for example, is composed of a hundred thousand or more tiny dots of different shades of gray or color, and the imaging system must be designed to ensure that they are small and numerous enough for a picture to be clinically useful-otherwise, it may appear blotchy, too noisy to be of value.

We have been discussing the factors that underlie the selection of a technology with suÐcient specificity and sensitivity for a given job, and the need for its images to be of adequate quality. Figure 7a, a candid portrait of young Nadine Wolbarst, nicely illustrates these ideas. So as not to produce just one more cute-kitten photo, I worked diligently to reveal the essence of her character by capturing, specifically, her glazed-over, catnip-deranged stare. The imaging equipment had to be sensitive enough, moreover, to perform in the challenging environment of a dimly lit suburban Washington den. Fortunately, disposable-camera film technology (with flash) was up to the demands, and the results exceeded my wildest artistic aspirations.

Figure 7b shows how the photo was made: Nadine was bathed in light, some of which was reflected toward the camera. The lens projected the pattern of light coming from her onto the film. The more light that struck a tiny area of film, the darker it would become when it was chemically developed later; thus, the camera produced a "negative." The process was then repeated, in effect, but this time with the negative (rather than Nadine herself) serving as the source of the incoming pattern of light; the negative from this negative was the positive" shown in figure 7a.

Technically speaking, figure 7a is not a bad picture. There is good contrast, easily picking up the shades of gray in Nadine's coat. The resolution is fine enough for us to make out her whiskers, and there is virtually no visual noise. Most important, at least with respect to the storage and transfer of information, the specificity, sensitivity, and overall visual quality are sufficient for you or me, the final link in the imaging chain, to determine that this is indeed a cat. And I, with access to certain additional data, can even assert with a fair degree of assurance which cat she usually is. If the film were underexposed or blurry, or in some other sense carried less information or more noise, then that would not necessarily be true.

Turning these thoughts back to medicine: no single technology can perform all imaging tasks well, so a physician must understand what the different types of medical images can reveal about a patient's condition. Only then can he or she choose the technology most likely to provide the essential, specific piece of information needed to address a given medical problem. Then it's up to the medical imaging staff to produce pictures of high enough quality to allow a reliable interpretation and diagnosis. And all of this has to be done with minimal risk to the patient and staff, and at an acceptable cost.

What Imaging Studies Reveal

Six general kinds of imaging are used routinely in modern diagnostic clinics: radiography, fluoroscopy (including studies that involve the computer), computed tomography, nuclear medicine, magnetic resonance imaging, and ultrasonography. The most familiar of these, of course, is radiography, the taking of X-ray films.

Radiography

As indicated above, medical images are generally produced by tracking the progress of suitable probes as they pass through the body. A beam of X rays consists of such probes (figure 8a). Think of an X-ray beam as a stream made up of vast numbers of small, discrete, particlelike bundles of energy, called photons. (Appendix A provides an elementary review of atoms and radiation.) X-ray photons propagate through space in straight lines and at the speed of light. Most important, they can collide with atoms and in this way be removed from the beam. In conventional radiography, a uniform, penetrating beam produced by an X-ray tube exposes a part of the body for a fraction of a second (figure 8b). Since the various tissues reduce the intensity of different areas of the beam by different amounts, an X-ray shadow is imprinted in the beam before it exits the patient. The shadow pattern that emerges is then captured on special photographic film. The more a bone or other tissue in the beam path absorbs or scatters X rays, the smaller the number of them that make it completely through to expose the film-and the clearer (less dark) the corresponding region of film will appear after it is developed.

X-ray films are most useful in locating and examining objects that have densities significantly greater or less than the surrounding soft tissues-as with bullets, bones, or lungs. X rays are also excellent for examining veins and arteries or parts of the gut if these areas can be filled with "contrast agent," such as certain compounds of iodine or barium, which soak up X rays particularly well. A tumor, unfortunately, presents more of a challenge. Because its density may be close to that of the surrounding healthy organ and muscle tissues, a cancer growth may give rise to little radiographic contrast, and so may be difficult (or impossible) to see directly on an X-ray film. Yet a tumor may reveal its presence by altering the appearance of an adjacent body (such as the wall of bowel that contains contrast agent) that can be visualized.

The inherently very high resolution of X-ray films enables them to provide critical details of fine structure, revealing hairline cracks in bone, for instance, and irregularities in narrow blood vessels enhanced with contrast agent.

Finally, it's easy to control most visual noise in film radiography, and it rarely causes difficulties, unless the film is under- or overexposed.

Conventional X-ray radiography is still the most common and least expensive way of obtaining diagnostic medical and dental images, and for many tasks it is perfectly adequate. But imaging departments have other options to choose from, as well, and in many situations one or more of them may offer a far better approach to a clinical problem.

Fluoroscopy

Fluoroscopy is radiography's first cousin. Here, the X rays that pass through and emerge from the patient do not immediately expose a film. Instead, they are projected onto the front face of an image intensifier (figure 9), an electronic vacuum-tube device that transforms a life-size pattern of X-ray shadows into a small, bright optical image. This visible image can be fed into a film camera; more commonly, it goes to a television (video) camera, where it is converted into an electrical signal and sent to a video monitor for live display. The image can be recorded on videotape for subsequent playback and further processing.

As with X-ray filming, fluoroscopy is most adept at distinguishing objects that differ significantly from soft tissue in density. Its major advantage is that it lets a physician watch bodily processes in "real time," as they happen-for example, the movement of barium contrast agent (given orally or by enema) past partial obstructions in the gut, or the passage of injected iodine-based compounds through constrictions in blood vessels.

By itself, fluoroscopy finds many routine applications in the clinic, but its considerable powers are extended even further when it is coupled to a computer. Digital subtraction angiography (DSA), in particular, is splendid for imaging arteries and veins-nothing else shows up on the screen but the arteries and veins (figure 10). The computer stores separate fluoroscopic images before and after contrast agent is injected into the patient's bloodstream. It then subtracts the first image from the second, point by point, and displays the difference between the two as a new image. This third, difference" image highlights those (and only those) places where the first and second images differ, that is, where blood vessels hold contrast agent; all the uninteresting, and easily confusing, background patterns are eliminated.

Computed Tomography (CT)

Conventional radiographic and fluoroscopic images are relatively straightforward and inexpensive to produce-even the smallest X-ray clinics have the equipment-and the trained eye can often derive more than enough from them for an accurate diagnosis. But the superimposed shadows from overlapping tissues sometimes obscure the critical details that a physician needs to see.

What is captured by film or fluoroscopy, and is thereby available for diagnosis, is the pattern of X rays transmitted through the body. The radiographic process is thus a kind of condensation, or deflation, of patient anatomy from the real world of three dimensions into a visual image in two. In the process, the shadows from an intricate three-dimensional structure, like a head, can be flattened into hopeless chaos on film, as in figure 5a. Although a radiologist may be able to perform near magic in detecting and interpreting slight irregularities amidst all the junk, in many cases there is simply too much visual confusion.

Digital subtraction angiography provides one path around that problem, but it works only for blood vessels. Computed tomography achieves the same end for a wide variety of organs, but it produces images quite differently. CT (called "see-tee" or "cat scanning") uses X rays, an elaborate radiation detection system, and a computer that carries out millions of calculations to construct the image of a thin, breadlike (transverse) slice of the patient's body (figure 11). By eliminating the interfering patterns that come from over- and underlying bones and organs, CT provides ample contrast among the various soft tissues, far better than standard radiography or fluoroscopy can do. So CT is routinely used for detailed studies of abdominal and pelvic organs, the lungs, and the brain. CT can image objects down to about 1 / 3 millimeter (one millimeter is roughly 1 / 25 inch); although that is not nearly as fine as the resolution in standard radiography, it is good enough for many diagnostic needs.

In the mid 1970s, CT gave physicians a whole new way of seeing, and the resulting effect on patient care has been incalculable. More recently, CT has had to face stiff competition from MRI, which provides clinical information that is usually comparable, and sometimes far superior. Still, when either modality can do a job just as well, the considerably lower costs of CT normally make it the appropriate choice.

Nuclear Medicine

Stolen cars and those bewildered wildebeests on Wild Kingdom can be located easily if they carry small radio transmitters. Nuclear medicine employs the same principle.

A nuclear medicine study makes use of a radioactive chemical substance, called a radiopharmaceutical, that displays two essential characteristics: it concentrates within a particular site in the body, an organ or tissue of interest, and it emits gamma rays. It thereby becomes a localized transmitter of gamma radiation. Radioactive atoms of the element technetium, for example, can be attached to microscopic clumps of molecules of a certain protein. When injected into the bloodstream, these radioactively tagged chunks become temporarily stuck in the narrowest blood vessels of the lung. Then, just as a red-hot poker glows in a dark room, a lung containing the technetium-labeled clumps will "glow" gamma rays. The gamma rays coming from the lung can then be detected and processed by a gamma camera and made into an image, as in figure 12.

A nuclear medicine image has relatively low spatial resolution; all you can see is the rough shape and size of the organ or tissue of interest. But if a part of the organ fails to take up the radioactive material, is missing, or is eclipsed by abnormal overlying tissues, then the corresponding region of the image will appear dark. Conversely, any part of the organ that takes up an excess of radiopharmaceutical will look unusually bright on the display. So a nuclear medicine image provides information principally on the physiology and pathology of an organ-what big parts of it are doing, or doing wrong-not the details of its anatomy.

Nuclear medicine began in the late 1930s when radioactive iodine was employed to investigate thyroid disease, and now it can contribute valuable information on nearly every organ. It has also worked well in the imaging of some tumors, providing an especially sensitive test of the spread of cancer from one organ to another. For cardiac patients, it furnishes quantitative assessments of the heart's capacity to pump blood.

One area where standard nuclear medicine has fallen largely out of favor (having been displaced by CT) is in brain imaging. Ironically, that's where positron emission tomography (PET) made its first important inroads. PET is a highly specialized form of nuclear medicine that utilizes a few unusual and diÐcult-to-produce atomic nuclei known as positron emitters in its radiopharmaceuticals, and some very complex and expensive imaging equipment. PET studies are particularly intriguing to neuroscientists and psychiatrists, since they can sometimes reveal the part of the brain where neural activity changes when certain mental processes are occurring. Figure 13a is a typical transverse-slice PET scan, at about the same level within the subject's head as figure 11, showing the distribution of water that has been labeled with a positron-emitting isotope of oxygen. One can stack many such slices to create a corresponding three-dimensional PET image. Figure 13b, for example, shows the differences that occur in such a three-dimensional map when a subject begins focusing on a specified task. The bright areas indicate changes in blood flow in the regions where the nerve cells are most active-in other words, not only is a new bunch of neurons starting to think deep thoughts, it is even managing to direct the circulatory system to divert extra jolts of fuel and oxygen there. To provide anatomical landmarks, this PET "difference" image has been superimposed on a picture of the brain's surface obtained separately with MRI.

Simple PET scans and task-related PET difference-images for people suffering from schizophrenia and other mental diseases look quite different from the normal. It may be diÐcult or impossible to explain what really causes such distinctions in terms of particular networks of nerve cells and blood vessels, but the images may still be diagnostically useful. Fortunately, you don't always have to understand a medical phenomenon completely to recognize it and deal with it effectively.

Long an important research tool in neurology, PET has recently been finding clinical applications, too, such as in assessing the suitability of a heart for bypass surgery and in searching for tumors.

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) not only reveals the structural details of the various organs, as does CT, but it also provides information on their physiological status and pathologies, as does nuclear medicine. And with MRI, there is no radiation risk to the patient, since no X-ray or gamma-ray energy is involved. Instead, MRI uses magnetic fields and radio waves to probe the (nonradioactive) nuclei of hydrogen atoms occurring naturally in the water molecules within and around cells.

Imagine a compass needle (which itself is a tiny bar magnet) aligned comfortably along the Earth's magnetic field. Now, in your mind's eye, twist it through 180 degrees, so that it points south, and then release it. It will flop back over, oscillate a few times with diminishing swings, and eventually come to rest pointing north again (figure 14a). The amount of time this settling-down process takes is known as the relaxation time.

Somewhat similar processes can occur with the magnetic nuclei of atoms in the cells of your body in a strong magnetic field, and it is these that MRI uses for imaging. Here is the central concept underlying how magnetic resonance imaging works: MRI produces a map of variations in the relaxation times of the hydrogen nuclei in the water molecules of tissues. But those relaxation times depend on the types of tissues involved, and even on their state of health. So by sensing slight differences in proton relaxation in physiologically different tissues, MRI can generate an image that distinguishes among them.

This may require a little more explaining. The nucleus of a hydrogen atom is a single proton, which behaves in some ways like a tiny magnet. The positively charged proton seems to be spinning rapidly and, as with any other moving electrical charge (such as current in a wire), it creates its own small magnetic field, rather like that of the compass needle. You can think of the "needle" as pointing along the proton's axis of rotation. So when a patient lies in the magnetic field of an MRI device, many of the hydrogen nuclei of the water molecules in the tissues will end up with their spin axes lying parallel to the field.

It is then possible, by beaming in radio waves of the correct frequency, to make a few of these protons flip over and point in the opposite (i.e., the wrong) direction instead. Immediately, however, protons will begin to undergo a kind of spontaneous relaxation analogous to (but quite different from) that of our compass needle, in which they flop back into their more comfortable, normal alignment, with the spin axis lying back along the field. The average time it takes for a bunch of protons to return to their stable, equilibrium condition is called a spin relaxation time.

MRI creates images out of what are, in effect, measurements of spin relaxation times of hydrogen nuclei. That is, an MRI device maps out, point by point within a part of the body, variations in the proton relaxation times of the intracellular and other water within the tissues. There are two related, but distinct, clinically important spin relaxation processes that can occur in any tissue, with times called T1 and T2. These are sensitive to, and reflect, subtle aspects of the interactions of the water molecules with the various other molecules in the tissue; the characteristics of those other molecules, in turn, depend on the type and physiological status of the cells that contain them.

Images that emphasize spatial variations in T1 or T2 show how rapidly these two, somewhat different, relaxation processes take place everywhere in a slice or volume of tissue. And even where tissues seem exactly alike to X rays, so that CT offers no helpful information, one or both of these forms of MRI image may be able to distinguish clearly between different organs, and even between healthy and pathological forms of the same tissue type, with splendid clarity.

MRI works by way of physical processes totally unlike those of CT, but it produces pictures that are remarkably similar to CT scans (figure 14b). It displays a degree of contrast among soft tissues that is often much better than that of CT. And MRI images can reveal subtleties in the physiology of an organ, not just its anatomy. It can even provide high-resolution anatomic views that CT simply cannot generate, as with figure 14c. The only serious drawback of MRI is the high cost of acquisition and operation, though this is decreasing.

Ultrasound

Unlike the other imaging technologies we have discussed, ultrasonography does not involve electromagnetic radiation, such as gamma and X rays or light or radio waves. Ultrasound is a mechanical disturbance, rather, in which oscillations of 1 to 10 million hertz (Hz; cycles per second) travel through soft tissues and fluids. (Humans with excellent ears typically can hear in the range 20 to 20,000 Hz.) In ultrasound imaging, a narrow beam of pulses of very high-frequency sound energy is directed into the body and swept back and forth; pictures of organs, blood vessels, and other structures are created out of the waves reflected back (figure 15).

When a pulse of ultrasound passes through different tissues, echoes are produced at the boundaries that contain or separate them. The time of return for each echo is proportional to the depth within the patient of the interface that produced it. The echo's intensity (amplitude) depends on that, and also on the differences in the acoustical characteristics of the materials on the two sides of the interface. A computer untangles all the data on echo times and intensities, and from that constructs images that can bring into view the beating of a heart or the stirring of an unborn child.

In many situations, ultrasound can provide critical clinical information quickly and inexpensively-for example, distinguishing a fluid-filled cyst from a solid tumor in the abdomen or the breast. It can assist in diagnosing a wide range of diseases, such as pathological changes in the liver, thyroid, gall bladder, pancreas, kidneys, heart, and the blood vessels. Ultrasound even serves as a guide in carrying out invasive procedures, such as draining an abscess. As with MRI, there is no X radiation involved, and it appears that a properly performed ultrasound study poses little risk to the patient. That is one reason it is used extensively to provide information about pregnancies-most importantly on certain fetal abnormalities-and as a visual aid in performing amniocentesis. Largely because of its cost-effectiveness and safety, some insurance companies and HMOs favor ultrasound for many kinds of examinations, and its role in the clinic will doubtless continue to expand.

Endoscopy

Finally, perhaps the simplest way to look within the body is, well, just to look. Physicians had long used rigid tubes, mirrors, and candles or sunlight to peer down the throat and along other bodily passageways, but it was the introduction of fiber optics in the 1950s that transformed the endoscope into a powerful modern clinical tool for direct clinical inspection.

Light entering the tip of a flexible glass fiber, a tenth of a millimeter or so across, will reflect again and again off its interior surface, like a stone skipping on water, and can travel great distances with little loss of intensity. A bundle of thousands of fibers, with lenses at the two ends, can carry a clear and sharp optical image. A modern endoscope consists of such a bundle, along with another bunch of fibers that bring in bright light to illuminate the region being viewed. In addition, it may have channels to convey gases or liquids in or out, or perhaps even a mechanical device, such as a biopsy forceps, at its business end. Endoscopy can be employed wherever there is an opening into the body, whether naturally occurring or surgical, and the equipment has become highly specialized for various applications.

As will be apparent from the case studies in this book, endoscopy may pick up where other imaging techniques leave off. Fluoroscopy may indicate the presence of a tumor of the esophagus, for example, but an endoscope then obtains the biopsy needed for confirmation (figure 16). Likewise, after angiography reveals a partial blockage of a coronary artery, a high-power laser beam carried by a special fiber optic bundle of an endoscope may burn it away.

An area where endoscopy is receiving much attention is small-incision-hole surgery. The endoscope enters the body through a cut as little as one centimeter long, and it either brings in the necessary surgical instruments itself or guides their use. While such operations are much less traumatic to the patient and result in shorter hospital stays (both of which reduce cost), the procedures require additional training for the surgeon.

I hope these brief sketches of the principal imaging technologies have whetted your appetite for a deeper understanding of them. If you like to know at the outset of a book whodunit, or at least how it's done, you can sneak a peek at the table in the epilogue that summarizes the physical processes involved in creating the various kinds of images, and their respective strengths and weaknesses.

The Tradeoff: Benefits versus Risks and Costs

X rays radically transformed the practice of medicine, but early on they revealed their dark side as well. Even before the turn of the century, researchers were reporting cases of horrible tissue burns caused by the radiation, some of which were fatal. Under the standard conditions used in diagnosis now, exposures high enough to cause such gross tissue damage are completely preventable. Unfortunately, though, radiation burns are not the only problem.

Laboratory experiments on cells and animals and epidemiological studies of exposed populations indicate that even at dose levels far below those that cause burns, X rays can nonetheless occasionally (but do not necessarily) induce cancers and birth defects. There is, however, a major problem with those midrange dose-response studies; the doses for which radiogenic (radiation-produced) cancers have actually been observed may be less than the burn level, but they are still a good deal higher than those that a patient receives in nearly any modern radiological examination. In other words, the midrange dose-effect study results can tell us almost nothing with assurance about the cancer risks at the very low-dose levels encountered in most X-ray or gamma-ray imaging. And direct evidence on the real risks from diagnostic amounts of radiation simply does not exist. Fortunately, what we do have suggests that the risks from today's diagnostic procedures are extremely small.

There is a fundamental tradeoff between the expected (but not certain) medical benefits and the highly unlikely (but possible) hazards from an X-ray examination; so also for other imaging methodologies. Over the years, medicine has learned to strike a proper balance between the recognized diagnostic advantages of imaging and the very small, but nevertheless real, possibility that serious, radiation-caused cancers and other adverse health effects might ensue from any study.

In the next chapter I shall say more about ways to gauge the risks for X rays. For now, I simply repeat that the estimated likelihood of someone becoming ill from a medically indicated, properly conducted diagnostic procedure is very, very small. The risk should be compared, moreover, with the probably far greater risk from not having the procedure performed, in which case a patient's treatment may be considerably less than optimal. So when someone asks, "Is this X-ray exam completely safe?" a good answer is usually: "We don't know for certain, but it's surely much safer than not doing the examination."

But what about the high costs of providing state of the art clinical care to large numbers of people? Hundreds of billions of dollars, a significant part of the gross domestic product, are spent each year on health care. A sizable block of this goes for the purchase and operation of big-ticket, high-tech medical equipment. Certainly, we all want the best possible diagnostic capabilities for those we love and for ourselves, but health care providers have to think carefully about whether one or another particular piece of costly apparatus is truly essential for proper treatment.

And as a community, we have but a finite pot of money. Given this limitation, would we all be significantly better off with a new MRI machine in the local hospital? Or with guardrails and lighting on that awful road that runs along the cliff, or with more police, or with better-paid teachers in the schools? Clearly, the benefits and the costs of imaging and other medical technologies must be weighed against the pros and cons of addressing our other major needs-and only through such an ongoing process can health care providers and society as a whole make intelligent decisions on the development and use of those technologies.

Fortunately, one of the most valuable forms of imaging, and certainly the most common, is also among the least expensive: ordinary X-ray filming. It is also the easiest to describe. So let us begin our full-fare tour of an imaging center with a visit to the X-ray suite.

Table of Contents

Acknowledgments
Preface

1. From the Watching of Shadows
2. Shadows on X-Ray Film: Radiography / Mammography
3. Shadows on Television, Live: Fluoroscopy
4. Shadows in Computers: Going Digital
5. Slices of Life: Computed Tomography (CT)
6. Like Embers in the Dark: Nuclear Medicine
7. Shadows from Echoes: Ultrasound
8. A Watery Mirror: Magnetic Resonance Imaging (MRI)
9. Epilogue: Looking Forward

Appendix A. Protons, Photons, and All That:
A Brief Review of Atoms and Radiation
Appendix B. More about MRI:
T 1 and Proton Spin Relaxation

Notes
Suggestions for Further Reading
Index