Aug 28 2019
A new study published in Circulation: Arrhythmia and Electrophysiology, a journal of the American Heart Association, reports that doctors could soon apply artificial intelligence to electrocardiogram data to evaluate the overall health status.
Suraj Kapa MD, Assistant Professor of Medicine and Director for Augmented and Virtual Reality Innovation, Mayo Clinic in Rochester, Minnesota. (Image credit: Mayo Clinic)
An electrocardiogram, also called ECG or EKG, is a test for measuring the heart’s electrical activity. Although it is known that the gender and age of a patient could affect an EKG, scientists proposed that artificial intelligence could establish the gender of a patient and estimate their “physiologic age”—a measure of overall health status and body function distinct from chronological age.
EKG data of nearly 500,000 patients were used to train a kind of artificial intelligence called a convolutional neural network to find similarities between the input and output data. After being trained, the neural network was tested for accuracy on the data of 275,000 more patients by estimating the output when only input data was supplied.
The neural network determined the chronological age of a patient as higher if they experienced adverse health conditions like heart attack, coronary artery disease, and low ejection fraction, and lower age after experiencing few or no adverse events.
While physicians already consider whether a patient ‘appears [their] stated age’ as part of their baseline physical examination, the ability to more objectively and consistently assess this may impact healthcare on multiple levels.
Suraj Kapa, MD, Study Author, Assistant Professor of Medicine, Director for Augmented and Virtual Reality Innovation, Mayo Clinic
He continued, “Being able to more accurately assess overall health status may help doctors determine which patients they should examine further to determine if there are asymptomatic or currently silent diseases that could benefit from early diagnosis and intervention.”
He further added, “For people at large, an AI-enhanced electrocardiogram could better show there may be something going on such as a new health issue or comorbid condition that they were otherwise unaware of.”
The scientists found out that the artificial intelligence had the capability to determine the gender of a patient accurately 90% of the time and could ascertain a patient’s chronological age group with an accuracy of 72%.
This evidence—that we might be gleaning some sort of ‘physiologic age’—was certainly both surprising and exciting for its potential role in future outcomes research, and may foster a new area of science where we seek to better understand the biologic underpinnings of such a finding.
Suraj Kapa, MD, Study Author, Assistant Professor of Medicine, Director for Augmented and Virtual Reality Innovation, Mayo Clinic
Although the research could draw from a large sample size, all individuals in the research were patients, and EKGs were performed for another clinical indication. Future analyses with an overtly healthy population are required to revalidate the determination of the neural network. Moreover, in the study, gender was self-identified by patients and may not represent the gender of all individuals in the study.
The co-authors of the study are Zachi Attia, MSc; Paul A. Friedman, MD; Peter A. Noseworthy, MD; Francisco Lopez-Jimenez, MD, MSc; Dorothy Ladewig, BS; Gaurav Satam, MS, MBA; Patricia A. Pellikka, MD; Thomas M. Munger, MD; Samuel J. Asirvatham, MD; Christopher G. Scott, MS; and Rickey E. Carter, PhD.
This research was financially supported by institutional funds at Mayo Clinic for data collection and statistical analyses.