AI estimates an area’s obesity rate from space

Sep 03, 2018, 8:26 AM EDT
(Source: Tony Alter/flickr)
(Source: Tony Alter/flickr)

The shape of a population is linked to the physical environment it inhabits; if the neighborhood presents a sprawling, open locality with gyms, parks and swimming pools instead of fast food joints, busy roads and grocery stores, you would see fewer love handles around.

An artificial intelligence program has been trained to spot the obesity rate of an area by analyzing its satellite imagery, a model that’s relatively much easier, quicker and economical than extensively surveying a neighborhood, reports Science.

The researchers, from the University of Washington, admit that such “eye-in-the-sky analyses” aren’t 100 percent accurate, adding that the approach can be handy in tracking obesity risk on large scale, notes Science Alert.