A selfie that can provide a lot of information-how can high-tech algorithms help those with serious diseases
Digital portrait (photo wave break media ltd, dreamstime.com)
An application developed with “Deep Learning” techniques can analyze the faces of cancer patients or other serious diseases and provide information that doctors use to choose how aggressive or “gentle” treatment is.
The researchers used a “deep learning” algorithm to measure the biological age of the subjects and found that the traits of cancer patients seemed, on average, about five years older than their age in the bulletin. The study was quoted by LiveScience and Science Alert.
Faceage estimates were also correlated with survival after treatment: the older a person seemed, regardless of her chronological age, the lower the chances of living. On the other hand, chronological age was not a good indicator of survival in cancer patients.
The study authors claim that the technology called Faceage could help doctors decide who can tolerate harsh treatments and who would need a milder approach.
The same logic could guide and complete the decisions on cardiac surgery, hip prostheses or life care.
There is more and more evidence that people grow old at different rhythms, depending on genes, stress, physical activity and tobacco and alcohol consumption. Although expensive genetic tests can show how DNA degrades over time, Faceage promises a similar perspective, using only a photo.
The model has been trained on 58,000 portraits of adults over 60, extracted from public data sets.
It was then tested on 6,196 cancer patients treated in the US and the Netherlands, using photos taken just before radiotherapy. Cancer patients seemed, on average, 4.79 years older biologically, than their real age.
The research in The Lancet has the title: Faceage, A Deep Learning System to estimated Biological Age from Photographs to Improve Prognostication: A model Development and Validation Study
Photo source: dreamstime.com
