November 13, 2019
The “die” has been cast—and an AI model knows it. It seems like something out of a thriller or a science fiction movie: Researchers from Pennsylvania healthcare provider Geisinger have trained an artificial intelligence model to predict which patients are at a high risk of dying within the next year, New Scientist reports.
The researchers tasked an AI model with examining the results of 1.77 million electrocardiogram tests conducted on nearly 400,000 participants in order to predict who was at a high risk of dying within the next year. The goal: To detect patterns that could indicate future cardiac problems, including heart attacks and atrial fibrillation.
As a control, the research team ran two versions of the AI: In one, the algorithm was given only raw ECG data, which measures voltage over time. In the other, it was fed ECG data in combination with patient age and sex.
The results were impressive (and a little scary). The AI model performed better than existing methods, according to the researchers, at distinguishing between patients who would die within a year and those who would survive.
“No matter what, the voltage-based model was always better than any model you could build out of things that we already measure from an ECG,” Brandon Fornwalt, lead researcher of the study at Geisinger, told New Scientist.
The model even detected heart problems in patients who were previously cleared by cardiologists.
“That finding suggests that the model is seeing things that humans probably can’t see, or at least that we just ignore and think are normal,” Fornwalt added. “AI can potentially teach us things that we’ve been maybe misinterpreting for decades.”
Research contact: @GeisingerHealth