Google AI can predict heart disease by scanning your eyes

Google’s ambition seems to continue to expand, the company’s machine learning assists are looking at cardiovascular disease.

In a paper published in Nature’s Biomedical Engineering magazine, Verily, the life sciences subsidiary of the search giant, described a method that predicts the risk of heart disease by scanning the posterior wall of the eye.

When you take pictures with a microscope and a camera, the part of your eye called “fundus” tells the doctor about the patient’s age, blood pressure, cholesterol levels, and whether they smoke – and doctors know if the patient will Possible heart disease. Using data from nearly 300,000 patients, Verily’s machine learning system was able to visualize retina images from two candidates, one of whom developed the cardiovascular disease over the next five years, while another No, the system’s prediction accuracy rate of 70%. This accuracy is as good as that currently being put into clinical use and requires blood testing methods.

However, the researchers also acknowledge that some limitations need to be overcome before preparing for the golden age of the technology.

Nevertheless, these findings are promising and the work done by Verily will certainly accelerate the development of cardiovascular disease testing to a greater extent.

Source: theverge