Scientists Accept Trained An Ai To Topographic Point Obesity From Space

Well, if yous are as well as therefore big they tin come across yous from space...

From ScienceAlert:

Our cities accept a lot to respond for.

What does your neighbourhood await like? Are in that place gyms, parks, swimming pools? Or are yous surrounded past times busy roads, fast nutrient outlets, as well as convenience stores?

The respond looks increasingly important. Obesity is a complex wellness issue impacted past times a range of factors, i of which is the physical, urban environment nosotros alive in. Now, scientists accept used artificial word (AI) as well as satellite images of U.S.A. of America cities to map that link – inwards effect, detecting obesity from space.

"We suggest a method for comprehensively assessing the association betwixt adult obesity prevalence as well as the built environment that involves extracting neighbourhood physical features from high-resolution satellite imagery," the squad explains inwards a new paper.

The researchers, from the University of Washington, fed some 150,000 high-resolution satellite images sourced from Google Maps into a convolutional neural network (CNN) – a sort of AI that uses deep learning to independently analyse as well as position patterns inside the dataset.

The information covered 1,695 census tracts inwards vi unlike cities: Bellevue, Seattle, Tacoma, Los Angeles, Memphis, as well as San Antonio.

The neural network the squad used inwards this illustration was already pre-trained on unopen to 1.2 i one 1000 thousand images – sense that helped it analyse the built surroundings across the cities, identifying features such every bit roads, buildings, trees, water, as well as land.

In addition, the researchers used estimates of obesity prevalence from the 500 Cities projection to practice a model that assessed the association betwixt those features (plus points of involvement similar gas stations, shopping malls, parks, as well as pet stores) as well as obesity prevalence inwards the studied areas.

It's non the laid about fourth dimension scientists accept done something similar this, precisely the researchers nation their technique is the most comprehensive endeavour yet.

According to their results, features of the built surroundings explained most two-thirds (64.8 percent) of the variation inwards obesity prevalence across all census tracts studied, although the success degree varied betwixt cities (with the highest beingness 73.3 pct inwards Memphis)....MORE

They don't "explain" they correlate, jeez.