Sniffing out cancer

When that chemical sticks to the DNA, an electric current flowing through the nanotube changes slightly, registering the difference. Team members are using molecular modeling to try to determine why a given strand reacts to a particular chemical, so they can custom-design them.

"Our hope is that by changing the sequence we can get sensors that respond differently to chemicals they are exposed to," Johnson said. The team has identified a dozen or so distinct chemical sensors so far with a goal of 100, he said, and has begun talking with potential investors.

If all goes as hoped, he said, a device might reach the market in six to 10 years. But the first versions would likely be for homeland security - not medicine, which requires far greater accuracy and presents big regulatory hurdles.

Plus, biopsy will almost certainly remain the definitive diagnostic tool for cancer for many years, said David J. Leffell, a professor of dermatology and surgery at the Yale School of Medicine.

Still, he was intrigued.

A device of the type that was described to him by a reporter theoretically could be used for broad-based population screening, Leffell said. And while basal cell carcinoma is more a destroyer of lifestyles than of lives, early detection raises a possibility of treatment with topical cream rather than surgical removal.

Leffell said the concept might also be more significant if it applied to melanoma, a less common but far more deadly form of skin cancer that is very treatable when detected before it spreads to the lymph nodes.

In fact, scientists at Monell will soon begin studying patients to develop an odor profile for melanoma. A device sensitive enough to sniff out cancer presumably could also find a number of other medical conditions, such as young children's amino acid disorders, whose smells are well-known.

Gelperin hopes his work will lead to "a new generation of electronic olfactory systems" for a wide range of uses. He wants an artificial nose that - like ours - can pull an odor's chemical signals "out of dozens and dozens of other signals that are floating around," a nose that can be trained to work when it "doesn't know in advance what it is going to encounter."

That sounds a lot like the tricorder of the Starship Enterprise. What, exactly, was the fictional gadget capable of?

Recalled George Preti, a Monell organic chemist who directed the work on the odor profile of skin cancer and also is part of the team developing nanosensors:

"It appeared to do everything."