Google AI scans eye images to predict cardiovascular problems

An artificial intelligence algorithm developed by Google Research may be able to predict a patient's risk of heart attack or stroke, according to a study published in Nature Biomedical Engineering Feb. 19.

The algorithm relies on deep learning, an advanced form of AI in which a computer learns over time, rather than having to be programmed like typical software. Software developers train a deep learning model by feeding it relevant data from diverse information sources.

For the cardiovascular disease project, an eight-person Google Research team trained deep learning models on retinal fundus images — photographs of the interior lining of patients' eyes — and medical data from 284,335 patients.

Researchers used two independent datasets of 12,026 and 999 patients to validate whether the algorithm would be able to identify cardiovascular risk factors based on associations between colors, patterns and shapes present in the retinal images.

The algorithm was able to identify various cardiovascular risk factors using the retinal images, including genetic factors like sex and lifestyle factors like high blood pressure. As one example, the researchers noted the algorithm was able to accurately distinguish between retinal images of a smoker and non-smoker 71 percent of the time.

The algorithm was also able to predict a patient's risk of experiencing a major cardiac event, such as a heart attack or stroke, within five years.

In a recent Google Research blog post, study author and product manager at the Google Brain Team Lily Peng, MD, PhD, wrote, "This performance approaches the accuracy of other CV risk calculators that require a blood draw to measure cholesterol."

To build on its findings, the Google Research team plans to continue developing and testing its algorithm on larger datasets.

More articles on artificial intelligence:
49% of organizations can't deploy AI projects because of data challenges, survey finds
Israeli company uses AI to predict which diabetes patients will suffer kidney damage: 3 things to know
Paige.AI launches with $25M in funding, data license agreement with Memorial Sloan Kettering

© Copyright ASC COMMUNICATIONS 2021. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.


Featured Whitepapers

Featured Webinars