Data mining of electrocardiogram histories is a key component of a new tool developed by a group of university and hospital researchers to better predict the risk of death in patients who have had a heart attack.
EKG data mining helps assess death risk for heart-attack patients: study
Results of the researchers' study and details of their tool to analyze patients' risk of death after a heart attack are published in the Sept. 28 edition of Science Translational Medicine.
Researchers from the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory (Cambridge), the University of Michigan (Ann Arbor), and Brigham and Women's Hospital and Harvard Medical School, both in Boston, used data mining and machine-learning techniques to analyze EKGs for 4,557 heart-attack patients.
"It doesn't require acquiring any data that's not already routinely acquired," said John Guttag, a MIT professor and one of the study's principal investigators, in an MIT news release. "It imposes no extra labor, no extra financial burden; it just takes data that they gather anyway and analyzes it in a different way.”
The tool identifies three new risk factors from a patient's EKG recordings. The three new metrics—computational biomarkers, in the researchers' terms—are morphologic variability, heart rate motifs and symbolic mismatch.
The metrics can provide information about heart defects or abnormalities, which, according to the researchers, are often overlooked during routine examinations. The study showed a "strong correlation between the three biomarkers and cardiovascular death over the two-year period following a heart attack," according to the release.
Send us a letter
Have an opinion about this story? Click here to submit a Letter to the Editor, and we may publish it in print.