Google’s AI algorithm now capable of predicting heart disease by scanning your eyes

Verily is a health-tech subsidiary of Google dedicated to developing tools that helps in collecting and organizing health data in order to come up with a platform that uses the information for health care management and provide interventions where necessary. The company recently discovered a way to use machine learning to predict the risk of heart disease for individuals.

A paper published in the Nature journal Biomedical Engineering reveals the approach being used to assess the risk of heart diseases which is by scanning the back of a person’s eye.

Some part of the paper which has been titled “Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning” reads:

The current standard-of-care for the screening of cardiovascular disease risk requires a variety of variables derived from the patient’s history and blood samples, such as age, gender, smoking status, blood pressure, body mass index (BMI), glucose and cholesterol levels. Most cardiovascular risk calculators use some combination of these parameters to identify patients at risk.

The search giant’s health company Verily goes ahead to state the results of applying deep learning.

Our results indicate that the application of deep learning to retinal fundus images alone can be used to predict multiple cardiovas-cular risk factors, including age, gender and SBP. That these risk factors are core components used in multiple cardiovascular risk calculators indicates that our model can potentially predict cardio-vascular risk directly.

The algorithm has being trained using dataset of approximately 300,000 patients. Google’s algorithm was able to correctly identify 70% of the time when presented with retinal images of two patients – one of whom suffered a cardiovascular event in the following five years, and one who didn’t.

Google’s researcher says there are still few limitations. Hopefully, there will be improvements before it becomes widely available for use as a method. We just have to wait patiently as the discovery is promising.

Via: The Verge


Leave a Reply