Using artificial intelligence, scientists at Google’s Health Tech subsidiary, Verily, can predict the risk of heart attack or stroke by looking at images of a patient’s retina. They claim that they can predict with 70% reliability which individuals could suffer a major heart attack over a five-year period.
This study, which has a cohort of 284,335 patients and was validated on the basis of two independent sets of data recorded for 12,026 and 999 patients respectively, has shown that rapid diagnosis is feasible without collecting any samples (e.g. for blood tests).
According to Doctor Lily Peng, the medical doctor leading the research, “This discovery will be finely tuned in the near future, to boost reliability and ease of application”.
If carried out as part of a routine health check, these scientific advances could help to evaluate and manage risks to patients in a rapid, non-invasive, cost-effective manner.
Top santé (21/02/2018) ; Europe 1(20/02/2018); Techxplore by Edward C. Baig (19/02/2018)