Google AI can predict heart disease risk from eye images

Joy Montgomery
February 20, 2018

Looking at deep-learning models on data from more than a quarter of a million patients, scientists were able to predict the cardiovascular risk factors that were not previously thought to be present.

Google scientists are presenting research on an algorithm that can accurately assess a patient's heart attack risk by examining retinal fundus images.

"Our approach uses deep learning to draw connections between changes in the human anatomy and disease, akin to how doctors learn to associate signs and symptoms with the diagnosis of a new disease", Peng said.

Veirly's software can also look at data such as individual's age, blood pressure, and whether or not they smoke before it can predict outcomes. According to the team, they were able to quantify the association between the retinal vessels and cardiovascular risks identified by researchers from previous medical studies.

Chelsea's Willian told Hull's David Meyler he wouldn't score penalty
Harry Kane, on the other hand, has again proved untouchable at Tottenham this season with a sterling score of 75/100. The game starts 0-0. "The game starts 0-0, and then we have to try to move the game in the right direction".

The scan takes place on the rear interior wall of the eye (AKA the fundus), where a large number of blood vessels that reflect the body's health exist. Peng said that they expanded their exercise and asked the deep learning model to predict whether a person was a smoker or what their blood pressure was based on retinal images. Rapid scans of the retina for example could be used as a screening tool for heart disease risk.

"In other words, it's not yet ready for clinical testing, but it's a promising start for non-invasive evaluation of cardiovascular health", Engadget concludes. This is only slightly worse than the commonly used SCORE method of predicting cardiovascular risk, which requires a blood test and makes correct predictions in the same test 72 percent of the time. These techniques helped the company to generate a heatmap which basically shows which pixels were the most important for a predicting a specific CV risk factor. "This discovery is particularly exciting because it suggests we might discover even more ways to diagnose health issues from retinal images", notes the search giant. Explaining how the algorithm is making its prediction gives doctor more confidence in the algorithm itself.

Google also used some attention techniques to find out how the algorithm was making its prediction. "However, we don't precisely know in a particular individual how these factors add up, so in some patients, we may perform sophisticated tests ... to help better stratify an individual's risk for having a cardiovascular event such as a heart attack or stroke", declared study co-author Dr. Michael McConnell, a medical researcher at Verily.

Other reports by PlayStation Move reviews

Discuss This Article