Conventional mammograms are simply black-and-white breast X-rays. Fatty tissue shows up as black, and everything else–glands, tumors, connective tissue–comes out white. Tumors stand out nicely against a backdrop of fat, but they’re hard to spot in the 40 percent of women with dense, fibrous breasts. The challenge is to find better ways of highlighting them. At the University of Toronto, biophysicist Martin Yaffe has spent more than a decade developing a technique known as digital mammography. Like any X-ray, it exploits the fact that tissues of different densities absorb different amounts of radiation. But whereas a film X-ray shows only gross variations, Yaffe’s device uses digital code to distinguish a range of subtle differences. Once an image is stored that way, a radiologist can manipulate it to highlight any worrisome element.

Several companies are now testing digital machines. They will cost roughly three times as much as the old ones, at least in the beginning, and no one has yet completed the large studies needed to prove that they’ll detect more tumors. But Yaffe’s device has performed well in pilot studies, and proponents believe the benefits of the new technique will justify the expense. Besides offering better resolution, digital breast images could be transmitted effortlessly from one computer to another, enabling surgeons and radiologists to consult over long distances. And instead of having to rely entirely on their eyes, specialists could use software programs to help analyze suspicious lesions.

X-ray isn’t the only way to spot a hidden tumor. Some radiologists have started using techniques like ultrasound and MRI (magnetic resonance imaging) to make dense breasts more scrutable. MRI is too costly to use for routine screening, but initial studies suggest that ultrasound–which uses high-frequency sound waves to map hidden structures–could become a useful adjunct to mammography. In one recent study Dr. Thomas Kolb, a New York City radiologist, offered ultrasound exams to 2,600 dense-breasted women who had already had routine mammograms and physical checkups. The extra tests revealed hidden masses in 195 women, 10 of whom turned out to have cancer. Finding those cancers cost $390,000 (each ultrasound exam runs about $150). But among Kolb’s patients, the extra testing improved the detection rate by 16 percent.

Better tests may improve detection, but the larger challenge is to make breast cancer less common. The disease may seem to strike randomly, but it’s a modern affliction, and many cases are theoretically preventable. Bad genes, though they place some women at extreme risk, account for only 5 to 10 percent of all cases. And genes don’t explain why the disease is five times as prevalent in North America as it is in Asia. When Asian women move to this country, their risk jumps accordingly. The reasons aren’t perfectly understood, but the likely culprits include a lack of exercise, obesity after menopause and a national diet that is rich in saturated fat and low in fruits, vegetables and fiber.

Diet may be important, but reproductive patterns have even stronger effects. Whenever a woman of reproductive age isn’t pregnant or nursing, her breast tissues get a monthly blast of estrogen. And estrogen, for all its virtues, can promote breast tumors. Anthropologists estimate that by menstruating earlier, reproducing later and spending fewer years nursing, modern Western women have more than doubled their lifetime estrogen exposure–and raised their lifetime breast-cancer risk to more than 100 times that of pre- agricultural women. Returning to the Stone Age isn’t an option, but modern medicine may yet devise a pharmaceutical equivalent. Researchers at the University of Southern California are testing drugs that could reduce women’s risk by mimicking the effects of pregnancy and lactation. That may seem an odd way to combat breast cancer, but it might accomplish what the most sophisticated screening never will.