LogitBoost algorithm predicts heart attacks, death with superhuman accuracy

A machine learning algorithm was able to predict death and heart attack from cardiac imaging data with more than 90 percent accuracy, according to the results of a study presented on May 12 at the International Conference on Nuclear Cardiology and Cardiac CT in Lisbon.

In the study, the LogitBoost algorithm was trained on the data of 950 patients with chest pain who, over the course of six years, had undergone various cardiac scans to detect coronary artery disease. The imaging results, alongside medical record data, produced 85 variables regarding factors like presence of coronary plaque, vessel narrowing, calcification, blood flow and basic clinical and demographic information.

The artificial intelligence system repeatedly analyzed those variables to learn how best to identify patterns in imaging data. Of the 24 heart attacks and 49 deaths that occurred within the six-year period, the algorithm retroactively predicted the outcomes with 90 percent accuracy.

"Doctors already collect a lot of information about patients — for example, those with chest pain. We found that machine learning can integrate these data and accurately predict individual risk," Luis Eduardo Juarez-Orozco, MD, PhD, author of the study and research fellow at Finland's Turku University Hospital, said in a statement. "This should allow us to personalize treatment and ultimately lead to better outcomes for patients."

More articles about AI:
AI can prevent hours of unnecessary testing for stable chest pain
AI retinal scanner that can detect Alzheimer's given FDA 'Breakthrough Device' status
Harvard Business Review: Look at AI as a system-wide update, not a point solution

© Copyright ASC COMMUNICATIONS 2020. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.


Featured Webinars

Featured Whitepapers