Mike Gillam, MD, FACEP, 23 July 2015
At Microsoft’s recent Build Conference, they revealed an experiment in machine learning – a web site that can guess the age and gender of faces in a photograph.
In a paper in Cell Research last month, Chinese researchers showed that facial analytics can indicate how quickly a person is aging better than other biological markers. They found that noses and mouths widen as people get older. The upper lip to nose distance increases and the distance between the eyes shrinks. The face ages of one group of forty-year-olds varied by as much as six years. By correlating their predictions with blood biochemical profiles, the researchers were able to show that their estimates provided a better guide of over-all health than age alone.
The New Yorker described that proxies to gauge a person’s biological age have been proposed – level of growth hormone, inflammatory markers, and bone density measurements – but none have been entirely reliable.
In the movie Gattaca, a future was envisioned where blood and genotype was used for the purposes of societal discrimination.
Yet, if phenotypic extraction could be coupled with genomic prediction, then the future of a Gattaca like society won’t require your blood – just your face.
—Mike Gillam, MD, FACEP 2015