Google AI is pushing to make a difference in the field of cardiovascular disease study by letting artificial intelligence predict potential heart attacks and strokes.
In a study published on Monday in the journal Nature Biomedical Engineering, Google AI researchers described how deep learning algorithms can be trained to predict heart disease symptoms by looking at retinal images.
“Traditionally, medical discoveries are made by observing associations, making hypotheses from them and then designing and running experiments to test the hypotheses. However, with medical images, observing and quantifying associations can often be difficult because of the wide variety of features, patterns, colors, values, and shapes that are present in real data,” the researchers wrote.@Google AI researchers made a breakthrough in cardiovascular disease study by teaching deep learning models to predict CV risk factors by looking at retinal images of patients. #ArtificialIntelligence #DeepLearningClick To Tweet
The team of researchers reportedly predicted cardiovascular risk factors that were “not previously thought to be present or quantifiable in retinal images” using the deep learning models that they trained on data from 284,335 patients. The information was then validated on two independent datasets of 12,026 and 999 patients.
Some of the CV risk factors that were successfully identified by the artificial intelligence system include:
- Smoking status
- Systolic blood pressure
- Major adverse cardiac events
“Here, we show that deep learning can extract new knowledge from retinal fundus images.”
Google AI to Conduct More Research
According to the Centers for Disease Control and Prevention, around 610,000 people die of heart disease in the United States every year, and 370,000 of these deaths are due to coronary heart disease (CHD). About 735,000 Americans also suffer from heart attacks annually.
With this alarming figure, the findings of Google AI is considered a breakthrough that could potentially save thousands of lives in the future.
However, despite the success of their experiment, Google said that more research has to be done. 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.
The Google AI researchers claim that they were able to predict whether a patient would experience a heart attack or other major cardiovascular episode within a five-year period. Their results indicate that their predictions are accurate 70% of the time.
“The caveat to this is that it’s early (and) we trained this on a small data set,” Lily Peng, an MD and lead researcher of the Google AI project, said. “We think that the accuracy of this prediction will go up a little bit more as we kind of get more comprehensive data. Discovering that we could do this is a good first step. But we need to validate.”
Google has allegedly been working on predicting eye diseases using artificial intelligence for a while now. 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. Google was surprised by the results and has pushed it further into predicting other CV risk factors.
Google AI’s method reportedly uses deep learning algorithms to create a so-called “heat map or graphical representation of data which revealed which pixels in an image.” Results are most significant when the algorithm was tasked with determining specific risk factors. For instance, the algorithm focuses on blood vessels when predicting blood pressure.
While Peng said that it might take years for their newly developed AI technology to be used in real-world medical applications, she’s optimistic that artificial intelligence can be applied in many areas of scientific discovery.
“I am very excited about what this means for discovery,” Peng said. “We hope researchers in other places will take what we have and build on it.”