Artificial intelligence can predict if, and when, a person could die of cardiac arrest.
We’ve seen AI used in a number of industries, and health care is no exception.
"What we set to do is to use imaging ... so we can predict risk of sudden cardiac death," said Dr. Natalia Trayanova, a professor of biomedical engineering and medicine at Johns Hopkins Medicine. She played a role in a new study looking at how AI can determine cardiac arrest. She is also the director of AI in health and medicine for AI-X Foundry, Johns Hopkins University’s new initiative bringing AI into all aspects of research.
Experts and doctors at Johns Hopkins have done multiple clinical trials looking at how AI can be used to help predict a number of diseases using MRI scans.
Thefirst study researchers released found that AI can predict the chance for sudden cardiac death in patients over a span of 10 years, and when it could happen.
Could AI be the key to detecting lung cancer years in advance?
Researchers and specialists teamed up to develop a new form of artificial intelligence that detects the deadliest type of cancer — lung cancer.
"One of the key important novelties in this approach is that we use raw images. We don't do anything to the images," Trayanova said.
Cardiac arrest causes about 300,000 to 450,000 deaths in the United States every year, according to the National Institute of Health.
While this new technology can help increase survival rates for patients, experts do have concerns over making sure the AI is ethical and equitable.
Trayanova said AI won't replace doctors, but it could be used to support decision making.
"I think that opens the door for so many things. There are so many clinical scans that we do on patients that are really underutilized and it shows how deep learning can figure out patterns and features that we don't know how to engineer, that we don't know about," she said.
Drones are being tested to deliver AEDs to cardiac arrest emergencies
Sweden is one of the first places to develop an AED drone delivery program, and over the span of four months, they made 11 deliveries.
