Diagnosing dementia early is both important and difficult. Carnegie Mellon University neuroscientist Anind Dey thinks hes found a better way.

By embedding motion-detecting sensors in domestic objects, such as coffee machines and pill boxes, Dey can make informational representations of people performing everyday tasks. Doctors can then look at that data for subtle changes suggesting mental decline.

Right now, the state of medicine being what it is, you have an occupational therapist come to your parents or grandparents house once every six months, said Dey, who presented his dwellSense platform Oct. 16 at the Wired Health Conference in New York City. For a diagnosis in which a few months can make a difference, thats not frequently enough.

So far, dwellSense has been tested with 14 elderly people in the Pittsburgh area. In addition to coffee-making and pill-taking, Deys group also tracked telephone-dialing. The information wasnt used for medical purposes, but to see if they tracked with traditional cognitive test results. Dey said they did.

They correlated very well with occupational therapist results, Dey said. It could catch problems more objectively, more quickly, than requiring an occupational therapist visit.

The next stage of the Robert Wood Johnson Foundation-funded project is collaboration with doctors who can give feedback on how best to present the data. We want to get them engaged in trying to figure out how to abstract this information to a higher level and improve the conversation with their patients, Dey said.

Go here to read the rest:
Sensors Embedded in Everyday Life Might Spot Dementia Early