Donald Patterson, a professor of informatics at the University of California, Irvine, says that the USF approach is more straightforward than those designed to monitor complex activities. "The more you get into the straight biological measurements . . . the easier it becomes," he says.

The USF technique relies on highly accurate RFID equipment. The ultra-wideband (UWB) chips used suffer less interference than do passive RFID chips and can send and receive signals through walls. The transmitters have a range of 600 feet and allow multiple people to be monitored even in a crowded room. The tags have batteries that last up to three years and accelerometers that put them into sleep mode when the user is motionless. According to Kearns, the entire system, including half a dozen tags, costs around $7,000 to implement.

Tanzeem Choudhury, an Intel researcher who uses RFID to gather social information, says that RFID technology is useful because it is so simple. "It's great they're showing a correlation with these [RFID tags]," she says.

Although walking patterns have been tied to dementia in previous studies, some experts question the approach. "There are a lot of factors that influence movement, and the disease in its very early stages is not a movement disease," says Robert Green, codirector of the Alzheimer's Disease Clinical and Research Program at Boston University. Green also points out that the USF researchers only looked for post-symptomatic dementia in their test.

However, Lisa D'Ambrosio, a research scientist at MIT's AgeLab, believes that the approach may be worth exploring. "It's a very interesting application of RFID technology," she says. "One of the trends in a lot of the neurology work is to move toward trying to diagnose Alzheimer's disease and cognitive impairment earlier."