Google AI looks at your eyes to predict heart disease risk

Desiree Steele
February 20, 2018

The work needs to be validated and repeated on more people before it gains broader acceptance, several outside physicians said. The algoritm was also able to infer a person's ethnicity, which is also a factor in cardiovascular related disease.

Harlan Krumholz of Yale University who was not involved in the research believes technology could soon "turbo-charge" diagnosis. "However, with medical images, observing and quantifying associations can be hard because of the wide variety of features, patterns, colors, values and shapes that are present in real images", researchers noted in a paper (PDF) published in the Nature journal Biomedical Engineering on Tuesday. An individual's blood pressure, age, smoking status has been earlier reported to be analysed by merely looking at retina scans and the pattern of blood vessels therein.

See, your eyes reveal a lot about your health.

Google and Verily's scientists analyse a medical dataset of almost 300,000 patients. Results are most significant when the algorithm was tasked with determining specific risk factors.

Sounds superficial? Well, it's a tested and proven method backed by Google's advanced artificial intelligence (AI) and machine learning in a project led by Alphabet's health-tech subsidiary Verily.

When presented with retinal images of two patients, one of whom suffered a cardiovascular event in the following five years, and one of whom did not, Google's algorithm was able to tell which was which 70 percent of the time.

Google indicates that its researchers will train the algorithms using larger datasets in the future.

Training deep-learning models on data from more than a quarter of a million patients, the scientists were able to predict the cardiovascular risk factors that were not previously thought to be present in retinal fundus images. This time, they also used a machine-learning technique, known as "soft attention", to help pinpoint which parts of the image were most instrumental in driving the algorithms' prediction. She added that explaining how the algorithm is making its prediction gives the doctor more confidence in the algorithm itself.

The idea that the hallmarks of disease could be detected through computational analysis has been alluring to engineers.

"To make this useful for patients, we will be seeking to understand the effects of interventions such as lifestyle changes or medications on our risk predictions and we will be generating new hypotheses and theories to test", Peng said. These diseases have remained the leading causes of death globally for the last 15 years, the organization noted.

Other reports by

Discuss This Article