Google AI can predict cardiovascular problems by looking at your eyes. Scientists from Google along with its health-tech subsidiary Verily successfully applied artificial intelligence and discovered a new way to predict the likelihood that a patient will suffer a heart attack or stroke through machine learning.
The method can assess the risk of cardiovascular problems such as heart attacks and strokes by examining images of the patient’s retina without any blood draws or other tests. But complete and careful testing of the method is required before using it in a clinical setting, according to a study of the finding published in the journal Nature Biomedical Engineering. Several outside physicians said suggests that the method need to be validated and repeated on more people.
This new approach makes it quicker and easier for doctors to assess the individual’s cardiovascular problems’ risk as it provides an efficient tool through which people can quickly and easily screen themselves for health risks.
To train the algorithm, Google used data from 284,335 patients which included eye scans as well as general medical data. Through deep learning analysis, neural networks were used to mine this information for patterns.
Lily Peng, MD, product manager and a lead on these efforts within Google AI said that they were able to predict cardiovascular risk from retinal images,
“For example, our algorithm could distinguish the retinal images of a smoker from that of a non-smoker 71 percent of the time, compared to a ~50 percent (i.e. random) accuracy by human experts,” Peng said.