firstname.lastname@example.org (Sophie Bateman)
A new algorithm has been developed that was able to correctly detect coronary artery disease with an 80% success rate by analysing the facial features of participants in a study.
Alopecia (hair loss), xanthelasmata (yellow eyelids) and arcus corneae (an opaque ring around the eye’s cornea) are among several clues in a person’s face that they could be suffering from poor cardiovascular health.
The algorithm needs more refinement before it can be rolled out as a useful diagnostic tool, but experts say it has the potential to “revolutionise medicine” — but there are significant ethical hurdles to clear first.
Between 2017 and 2019, researchers photographed 5,796 patients who had gone to hospital for Heart imaging procedures. Four images were taken of each patient — one full-frontal, two side profiles and one looking down at the top of the head.
It was also capable of correctly detecting patients without coronary artery disease 61% of the time.
“The algorithm had a moderate performance, and additional clinical information did not improve its performance, which means it could be used easily to predict potential Heart disease based on facial photos alone,” says Xiang-Yang Ji, one of the researchers working on the study.
“The cheek, forehead and nose contributed more information to the algorithm than other facial areas. However, we need to improve the specificity as a false positive rate of as much as 46% may cause anxiety and inconvenience to patients, as well as potentially overloading clinics with patients requiring unnecessary tests.”
“It is a step towards the development of a deep-learning-based tool that could be used to assess the risk of Heart disease, either in outpatient clinics or by means of patients taking ‘selfies’ to perform their own screening.”
Charalambos Antoniades and Christos Kotanidis, cardiovascular experts from the University of Oxford, say once the selfie-reading technology is accurate enough for mass use it could “revolutionise medicine as we know it”.
However they have also warned that these kinds of diagnostic tools will face major ethical challenges, such as if non-medical professionals are able to collect such private health information from photos of people.
“Such a technology may raise concerns about misuse of information for discriminatory purposes,” the pair wrote in an editorial about the algorithm.
“Unwanted dissemination of sensitive health record data, that can easily be extracted from a facial photo, renders technologies such as that discussed here a significant threat to personal data protection, potentially affecting insurance options.”
Peter Bannister, from the Institution of Engineering and Technology, also voiced his concern that the technology “could be maliciously applied on data in the public domain”, like a screenshot from an online conference call being used to deny someone certain health insurance in the future.