The health technology division of Alphabet Inc. (Google’s parent company) now has a new Artificial Intelligence algorithm. That can predict the possibility that a patient may suffer cardiac risks with almost the same precision as current methods. But this new method is based on scanning the back of the eye. So, that avoids the need to perform blood tests, obtaining the results more quickly.
New AI algorithm can predict heart risks by scanning eyes
To this end, training has been conducted through neural networks using medical data from almost 300,000 patients. That includes eyeball scans and general medical data. With which the algorithm itself has been able to extract patterns to associate them with possible signs of leading to risk. Cardiac The results of the predictions are somewhat worse than with conventional methods, but not much. Since the possibility of making correct predictions is 70% compared to 72% of traditional methods.
To reach up to this point has been taken into account those investigations that point to the blood vessels of the eyes can reveal some information about the patients themselves. Hence the new algorithm allows doctors to be more efficient. There is still a long way for the algorithm to be used in clinical settings, so there is still a period of maturation and improvement.
This algorithm of Artificial Intelligence opens the doors to health diagnostics without human intervention. But to reach this point there are still years, or even decades, to become a reality.
A new article has been published today about the work done on this algorithm in the journal Biomedical Engineering of the journal Nature, according to The Verge.