Implantable Dust-Sized Sensors to Monitor Health

A small speck of a sensor -- no wider than a human hair -- can detect light, pressure, temperature, and the presence of certain chemicals. The device, which uses flash-memory technology, the same memory used in some iPods, could eventually be employed in a wide variety of applications, including better drug-screening tests, the continuous monitoring of the health of organs and blood vessels, and as tiny sensors for detecting chemicals in the environment. The sensors, invented by Edwin Kan, professor of electrical and computer engineering at Cornell University, are now being developed for commercial use by Transonic Systems in Ithaca, NY. The first such application, which could be available in five years, will probably be for studying and monitoring blood flow, pressure, and temperature in the small blood vessels of lab animals, an important part of drug trials, says Transonic project engineer Bruce McKee. In such an application, the sensors would be implanted in the bloodstream of animals. In fact, because the sensors are so small and consume so little power, McKee says it's possible they could be permanently implanted in the human body, along with a radio for communications and a power source. Such an implant could continuously monitor the health of organs or levels of certain compounds. Or an implanted pressure sensor, he says, might be useful for monitoring the pressure inside the eyes of those with glaucoma, a leading cause of blindness that's associated with increased eye pressure. The sensor is based on the transistors in flash memory, the type of memory used in digital cameras, some portable electronics, and cell phones. Flash memory stores information by changing the electrical charge in a structure called a floating gate. Kan has added "sensing gates" to these transistors that, depending on the material they're made of, will selectively change their charge in response to pressure, heat, light, or the presence of certain molecules. This change in charge, in turn, alters the charge in the floating gate, which can then be detected the way information is normally retrieved from flash memory. Kan has built prototype sensors measuring 100 micrometers on a side, but says they could be made much smaller. Right now, the sensors communicate their information and receive power via wires. Adding a radio and power for a permanently implantable wireless device will increase the overall size to a couple of square millimeters, Kan says. The Cornell research is part of a growing effort to develop implantable sensors. One example already in production is a pill that once swallowed can take pictures of the intestine and send them wirelessly to a computer. Kan says his devices are much smaller and would require significantly less power. The flash-based system is a general platform that could be used for a wide variety of sensing applications, says Bradley Minch, professor of electrical and computer engineering currently at Olin College; while a professor at Cornell, he worked with Kan to develop an earlier version of the sensor. These sensors should also be easy to manufacture, he says, since they are similar to existing flash-memory devices. Joseph Paradiso, a professor in MIT's Media Lab who develops wireless sensor networks, says the approach is "intriguing." Given the high density possible with flash memory cells, he says it may be possible to make arrays of sensors for detecting various chemicals -- all on one chip. The sensors might be useful as part of sensor networks, he says, perhaps embedded in the "skin" of robots, for detecting things such as pressure and temperature in addition to chemicals. "Anything like this is great, and there are always applications that people come up with," Paradiso says. "Source":[ http://www.technologyreview.com/read_article.aspx?id=16993&ch=biztech]

---