Someone exhibits a warning sign for cancer–an odd-shaped mole, problematic changes in cells removed from an in-body surface–and the doctor’s worried.

Typically in such circumstances, a doctor will ask that a small sample of tissue be removed. The sample’s sent to the lab, and hours or sometimes days later, the results come back.

It’s a time-tested way to probe for cancer. But imagine the gains in both convenience and safety if the doctor could simply run a pen-sized scanner over the site and get results on the spot.

The approach is called a non-invasive biopsy, and MIT’s Peter So says that “in some ways, it’s the Holy Grail of bioimaging technology.” So, an assistant professor of mechanical engineering, is in all-out pursuit of this particular grail.

A Hong Kong native, So spent his high school years in Hawaii before coming to the U.S. mainland for college and graduate school. He’s a tinkerer whose grade-school exploits included building walkie-talkies and a model version of a railroad engine powered by real steam. (“It did work,” says the faculty member, “though I won’t claim that it worked great.”)

Peering under the skin

So studied physics in college and grad school, but never lost his taste for building things. Since the early ’90s, that interest has helped drive his efforts to build devices that let you peer into and through the outer layers of human tissue, probing for subtle hints of disease.

So’s approach reflects two facts about the human body. One is that certain types of light–notably red and infrared–can penetrate the skin and like tissues. (That’s why your cheeks glow red if you put the head of a lit flashlight in your mouth.) The second is that, in the right circumstances, we “glow.”

“As it turns out, we’re all fluorescent,” explains So. Not that our bodies light up like a policeman’s vest in a car’s headlights. But when you target human tissues with light of the right wavelength, certain of our native molecules fluoresce.

That wouldn’t help if the triggering light couldn’t reach those molecules. But though red and infrared light penetrate only about as far as this page is thick, that’s deep enough to potentially let doctors pinpoint many sub-surface tumors.
Cell shape and cancer

So’s approach is based on a technology called two-photon microscopy–a name that reflects the physics involved. And his goal is to use those almost infinitesimal dots of fluorescence in human cells to take a detailed peek under the surface of our tissues.

“The idea,” he explains, “is to look at the structure and the so-called morphology, or shape, of these cells and determine if they are cancerous.” In many cancers, he notes parenthetically, changes in cell shape are the best give-away that cancer is present.

The system, already tested in biopsy samples, yields spectacular images: colorful 3-D “pictures” of tissue cross-sections that the user can zoom in on, rotate, and even turn over on the computer screen.

So has yet to prove that such images can offer a surefire test for cancer. Working with Massachusetts General Hospital dermatologists, though, he’s currently addressing the problem. “We’d like to be able to judge cell morphology about as well as you can with a biopsy,” he says.

If the researchers succeed, as recent results suggest they will, it could ease the challenge for doctors worried about many types of problems. One obvious application is scanning worrisome moles. But the system could also give doctors a way to tell if they’ve completely excised an already-identified tumor, potentially minimizing the amount of healthy tissue that has to be cut out.

“If you’ve got a lesion on your arms or legs, that may not be a big deal,” says So, “but if it’s on the face, that could determine whether surgery is disfiguring or not.”

Scanning inner surfaces

So is also working with colleague Ian Hunter on systems designed to probe inner-body regions–the digestive tract, the cervix, the uterus–where tumors can form on or just below surfaces. “If you had an abnormal Pap smear,” notes So, referring to a standard screening test for cervical cancer risks, “you could use a two-photon system to scan the cervix for evidence of cancer.”

A key priority, meanwhile, is making systems that work fast. “In a clinical setting,” he notes, “it’s not feasible to ask a patient to sit for an hour while you collect the data.” Using innovations like a light-scattering mirror that spins up to 30,000 times a second, So has cut how long it takes to scan, say, an average-sized mole to just a few minutes.

For So, these efforts bring multiple rewards. One is that he gets to keep building things. In fact, he and his co-workers have finished or are well along on 10 different two-photon systems, with groups from as far away as California as recipients.

Also, there is working with physician colleagues. “The fact that MIT is in the middle of a major medical complex means the opportunities for collaboration are tremendous,” he notes.

And, too, he’s in a realm of science and technology where the chances to make a difference are real and enticing.

“In physics, there are mostly big questions,” notes the former physics major. “There are big questions in biology, too, but there are also smaller questions, and if we can answer some of those we can help society at the same time we are learning new things.”