Ninety years ago next month, William Konrad Roentgen, then professor of physics at the University of Wurzburg in Bavaria, accidently discovered a new kind of ray. He called it an "X-ray" because he didn't know what it was. But he did know that it could penetrate solids opaque to other lights and deposit images of internal structures on photographic film.

As scientists immediately began using the discovery to detect disease, the public's reaction was mixed. Within weeks, English entrepreneurs advertised "X-ray-proof underclothing -- especially made for the sensitive woman." Spiritualists, telepaths and clairvoyants said the new rays would provide long-sought facts to prove their claims. In the United States, moral brigades assembled to resist the destruction of all decency and privacy. Public-spirited citizens protested the use of X-rays to pry into people's minds and habits.

While no device can yet see into the mind, Roentgen's rays and the $2 billion-a-year industry they created can see inside the body with startling clarity and photograph everything from a tiny tumor in the brain to painful grains of calcium in the kidneys.

In the last decade and a half, there has been an explosion of diagnostic imaging technologies that rival the impact -- and fulfill the promise -- of Roentgen's original discovery.

Enhanced by computer technology, the two-dimensional X-ray pictures that have been the physician's basic imaging tool from the turn of the century to the 1960s have been replaced with three-dimensional images not just of bone, calcified tumors and dyes, but also of the soft tissue organs such as the heart, the brain and the liver. In addition to producing pictures of these organs, these techniques allow doctors to see anatomical changes such as infections, tumors or blockages and are even now beginning to measure an organ's metabolic activity.

"Twenty years ago, we thought we were seeing most everything with chest films, barium enemas and bone films," says Dr. Bernard Smith, associate chairman of the department of diagnostic radiology at Arlington Hospital. "Now we realize that there was a big new world of imaging that we were not seeing."

A patient entering a hospital's radiology department today encounters a host of diagnostic imaging devices, including magnetic resonance imagers (MRI), computerized axial tomographs (CAT or CT scans), positron emission tomographs (PET scans), ultrasound, digital subtraction angiography (DSA), nuclear medicine, angiography and, of course, the traditional X-ray.

"I can't think of another field in medicine where the front is moving ahead so impressively," says Dr. Seymour Perry, senior fellow at Georgetown University's Institute for Health Policy Analysis.

And the radiologists -- who until now "have been imagers and diagnosticians and have not done therapy," says Dr. Harold Coons, director of interventional radiology at Sharp Memorial Hospital in San Diego -- have begun to treat diseases. Guided by sophisticated images of internal organs produced with these new technologies, the radiologist can precisely place a needle into a deadly abscess to drain infecting bacteria or pluck stones from the kidney and gall bladder through tubes and even unblock clogged arteries.

This radiologic revolution "certainly has meant earlier diagnosis, better selection of treatment, less hospitalization, less radical surgery and, depending on the case, better survival," says Arlington Hospital's Smith.

Although the revolution in imaging technology has been exciting, it has raised serious questions about the cost of the equipment and its impact on the overall cost of health care.

The Public Health Service is now recommending that one of the newest of these technologies, magnetic resonance, be approved for use on Medicare patients. If this recommendation is approved, it could pave the way for introducing some of the most advanced medical technologies -- capable of producing enormous health improvements and generating huge costs -- into the nation's largest medical program.

X-rays are part of the same electromagnetic spectrum as visible light, but they can penetrate solids because they have, in a sense, more energy than visible light.

X-rays punch through the soft tissues of the body like a high-powered bullet through a wooden door. As they come out the other side, X-rays can be used to expose sensitive photographic film. If the X-ray passes through the body without interference, it strikes the X-ray film and turns it black. But if the X-ray beam is turned aside, deflected or reflected by dense tissue such as a bone or calcified tumor, then the X-ray beam is scattered and much of it fails to strike the film, leaving that portion of the film white or gray.

Consequently, the images left on the X-ray are actually the shadows of structures within the body.

The basic physics is the same no matter which kind of imaging machine -- CAT scanner, fluoroscope, angiographic imager -- is used.

For decades, radiologists have stared at the shadows created by X-ray beams to detect disease as subtle alterations in the internal anatomy. These kinds of studies, however, have limitations.

X-rays are able to distinguish only between air, fat, water and bone. They cannot differentiate between a tumor and an organ -- such as the liver -- because the tumor and the liver have essentially the same density. They also cram information about a three-dimensional object into a two-dimensional image, which causes distortions and often makes a diagnosis impossible.

Advancements in imaging technologies over the last 30 years sought solutions to these limitations.

The 1950s brought the development of image intensifiers -- electronic devices that capture the X-ray and convert it to a sharp, bright image on a screen. Intensifiers cut the radiation dose needed to get a picture and improved the quality of the image, but it was the computer that created the technological explosion in imaging equipment.

"Even the standard chest X-ray has been computerized," says Joseph N. Williams, president and chief executive officer of Picker International Inc., the second largest X-ray equipment manufacturer in the United States.

"Computers allow you to change the contrast [the intensities of black, whites and grays]," says General Electric's Bob Moliter, manager of the Medical Systems Group's government programs office in the District. "You can zero in on a region and magnify it. You can reverse the contrast to positive or negative images. You can superimpose consecutive scans -- one with contrast, one without. It permits you to drop out certain extraneous parts of the image and look only at the area of pathology or clinical interest."

And once the image is inside the computer, the data can be transmitted nearly anywhere -- as Picker will demonstrate next month, when it beams pictures from a magnetic resonance imager at the National Institutes of Health's Clinical Center in Bethesda to Chicago, where the Radiological Society of North America will hold its annual meeting.

But the most important application of computers began in 1973, when the first computerized axial tomograph, better known as a CAT scanner or simply as CT, was introduced in the United States. With CT, the computer synthesizes an image from a collection of fan-shaped, pencil-thin X-ray beams that are projected through the patient. The result is an image that looks like a cross-section slice of the organ scanned. The computer processes used in CT originally were created to beam pictures of planets back to Earth from various space probes.

"The major breakthrough in CT is that for the first time it [was possible] to get a cross-section of images of high quality without the superimposition of structures above or below the area you are interested in," says Wilfried Loeffler, technical manager of magnetic resonance and computerized tomography for Siemens Medical Systems Inc., a large equipment manufacturer.

"CT allowed us to see organs we never saw before, like the heart and the pancreas," says Dr. John Doppman, chief of diagnostic radiology at the NIH Clinical Center. "CT was the biggest advancement in imaging in 50 years."

From the computer systems developed for CAT came two new imaging systems: magnetic resonance imaging and positron emission tomography. Neither uses an external X-ray beam, but both handle the data in much the same way as CAT scanners and produce similar, but significantly different, images.

Magnetic resonance imagers, or MRI, "has a number of distinct advantages," says NIH's Doppman. "It is exciting, but extraordinarily complex."

This technology -- which was once called nuclear magnetic resonance, but leaders in the field dropped "nuclear" because it made patients think it was radioactive -- uses no X-rays or radiation at all. Instead, the images are created by placing the patient in a super-powerful magnet, about 100,000 times more powerful than the magnetic field of the earth. "There is no evidence that strong magnetic fields have any adverse biologic effects," says Doppman.

The magnet forces hydrogen atoms within the body -- which can be imagined as toy tops all spinning in different directions -- to line up.. By pulsing waves of a certain radio frequency through the patient, the spinning hydrogen atoms are knocked out of alignment with the magnetic field. When the radio waves are turned off, the spinning hydrogen atoms realign themselves with the magnetic field, giving off their own radio signals in the process. The hydrogen's radio waves are picked up by an antenna in the MRI unit and converted into images by the computer.

Because MRI is "seeing" freely moving hydrogen atoms, it is able to separate different tissues based on the concentration of these atoms. Fat, which has a high concentration of free hydrogen, can be separated from muscle, which has far fewer; tumors can be distinguished from nerve tissue, and bones do not show up at all because they contain little water or free hydrogen.

MRI produces images of tissues that cannot be seen even with CT -- views along the length of the spinal cord, for example -- and it already has been shown to more effectively diagnose neural diseases, such as multiple sclerosis, than any of the other imaging techniques.

Another modern imaging tool, positron emission tomography or PET, combines computer technology with radioactive particles to produce a three-dimensional look at the metabolic activity of, say, different regions of the brain.

The radioactive tracers used in PET have short half-lives, which means they exist for only a brief period. Half of the radioactive oxygen used in some PET studies is gone in two minutes. Half of the radioactive fluoride, which is currently used in the brain studies, disappears in less than two hours.

As these radioactive atoms decay into their stable, nonradioactive form, they emit positrons, which give off a characteristic signal that the PET scanner can detect.

PET "traces biological molecules in the body," says Dr. Steve M. Larson, chief of the department of nuclear medicine at NIH's Clinical Center, where the installation of a new PET system is nearly completed. "The basic importance is that we can measure the body's metabolism" for the first time.

Instead of seeing an anatomical image of the inside of the brain, the PET measures its activity. PET is being studied as a way of diagnosing epilepsy, stroke and Parkinson's disease, tracking blood flow in the body and following the progress of tumors.

But has this explosion of technology made a difference in the way patients are treated?

"Patient care is improving," says NIH's Doppman. "We are making diagnoses that we did not make 20 years ago."

But no one can really say how much of a difference these systems are making in the way patients are treated or in how long they live, says Dr. John Marshall, director of the National Center for Health Resources Administration and Health Care Technology Assessment.

"We are trying to figure out what the explosion of technology means," Marshall says. "Can you quantify how much good it does? Theoretically, but it is hard." Good studies just don't seem to exist.

Even choosing the best technique to make a diagnosis is not always easy. During a meeting of the American College of Nuclear Medicine in the District this fall, members of panel of physicians each argued that their particular way of seeing the heart -- fast CT, ultrasound, or thallium scanning -- was the best approach.

The choice of a particular imaging system is important, since imaging systems themselves have begun to play a role in therapy.

The central theme in this new field is the ability to use sophisticated imaging technology to selectively place a needle or a catheter through the skin to reach some problem: an infection-filled abscess, an obstructed kidney or gall bladder, a blockage in a common bile duct, and suspected tumors within the pelvis, abdomen and chest that need to be biopsied.

"It is remote control surgery through small tubes," says Dr. Harold Coons, of Sharp Memorial Hospital in San Diego and an outspoken proponent of this new field. "It is a very different role for radiologist. Now we are fixing patients rather than just diagnosing them."

For example, "draining abscesses is a very common procedure. These occur after surgery," Coons says. "To drain a small abscess deep within the body, we identify it with CT, and, under CT guidance, we put the needle in the abscess. Then take the patient to fluoroscopy, fill the abscess with dye to image it," then slide larger and larger catheters into the infected region so it can be drained.

"We can save a second surgery" with this technique, Coons says. "Surgeons have not liked it. We do more than 1,000 procedures of all kinds" each year at Sharp, which is a 400-bed hospital.

While all this has happened because Roentgen accidentally discovery X-rays nearly 90 years ago, it has been the application of computers that intensified and clarified the images of the body. One journalist might well have been writing of CT or PET or MRI today when he penned in 1896: "Diagnosis, long a painfully uncertain science, has received an unexpected and wonderful assistant . . ."

And the next step, predicts Williams, president of Picker International: Artificial intelligence now being developed for the next generation of computers will be hooked into future imaging equipment so that the machine itself will make decision about the best way to diagnose the disease.

"We are trying to introduce it," Williams says, "slowly but surely, with our MRI work."