Early Research Suggests Promising Use of AI to Predict Heart Attack or Stroke Risk Using a Single Chest X-Ray

CNN
—

Early research suggests a promising use of artificial intelligence to predict the 10-year risk of death from heart attack or stroke from a single chest X-ray.

Preliminary results were presented Tuesday at the annual meeting of the Radiological Society of North America. The research is in the final draft stage and has not been submitted for publication in a medical journal.

The researchers used nearly 150,000 chest X-rays to train an artificial intelligence program to identify patterns in the images associated with the risk of major cardiovascular events. They tested the program on a separate group of around 11,000 people and found a « significant association » between the level of risk predicted by the AI ​​and the actual occurrence of a major cardiovascular disease event.

The clinical standard for cardiovascular disease risk analysis is the Atherosclerotic Cardiovascular Disease Risk Score (ASCVD), a calculator that weights various patient data points that have been shown to have a strong association with adverse cardiovascular events, including including age, blood pressure and smoking history.

Statins are recommended for people with a 10-year risk of 7.5% or more. The AI ​​model uses the same risk thresholds as the established risk calculator, and early findings suggest it works just as well.

“We have known for a long time that X-rays capture information beyond traditional diagnostic results, but we did not use this data because we did not have robust and reliable methods,” said researcher Dr Jakob Weiss. Principal and Affiliate Radiologist. with Massachusetts General Hospital and the AI ​​in Medicine program at Harvard Medical School’s Brigham and Women’s Hospital.

Sometimes the AI ​​results line up with a traditional radiology reading, but other times they pick up things that may have been missed, he said.

« Part of these are anatomical changes that we would also detect with the naked eye and that make physiological sense. Let’s say there is an increase in blood pressure or heart failure – these are findings that we can also detect on a normal chest X-ray But I think a lot of the captured or extracted information is embedded somewhere in the scan, but we can’t understand the meaning of that as traditionally trained radiologists at this time said Weiss.

“It has this black box character,” he said, which can sometimes make it difficult to communicate the risk to patients without an explanation to identify.

Dr. Donald Lloyd-Jones, chair of preventive medicine at Northwestern University’s Feinberg School of Medicine and past president of the American Heart Association, was co-chair of the risk assessment committee when the ASCVD risk calculator was created in 2013 and a key player in 2018 when the guidelines were updated to highlight the relationship between risk score and personal medical history.

He has not been involved in new AI research, but says it is important to move the field forward.

« This is exactly the type of application that artificial intelligence is best for, » he said. “So we have to keep doing things like this to really understand if we can find, in particular, patients who would otherwise fall through the cracks. I think that’s where it can be most useful.

But collecting all the patient data points that go into the established risk calculator is still essential, because they are actionable. And whether the risk is calculated using a statistical formula or an AI model, the most successful outcomes will always require personalized patient assessments.

“We don’t cure smoking with a chest X-ray. We actually have to work with the patient to find ways to get them to quit smoking,” Lloyd-Jones said. “The risk calculator is one part of the risk assessment, but it’s not the only part. It’s a process that involves both the patient and the doctor in a discussion about the patient’s risk and how well we think a statin would help them.

For their research, Weiss and his co-authors trained the AI ​​using chest X-rays of participants in the National Cancer Institute’s Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. It was tested on people who had routine outpatient chest X-rays at Mass General Brigham and were potentially eligible for statin therapy, with an average age of 60.

Further research, including a randomized controlled trial, is needed to validate the deep learning model.