AI: Artificial Intelligence Helps Detect Early Heart Disease - Mayo Researchers

Artificial Intelligence, utilized in conjunction with EKGs, could be the key to detecting early heart problems, a new study from mayo Clinic found.
Artificial Intelligence

An AI system can effectively detect an early, pre-symptomatic stage of heart disease using simple electrocardiogram data. Applying artificial intelligence (AI) to a widely available, inexpensive test — the Electrocardiogram (EKG) — results in a straightforward, affordable early indicator of a precursor to heart failure, scientists say.

Artificial Intelligence and Health

An innovative new way to analyze electrocardiogram (ECG) results using artificial intelligence might facilitate doctors quickly and cheaply identify patients most at risk of heart failure.

A new study from researchers at the mayo Clinic demonstrates a newly developed AI system that's able to detect the presence of left ventricular dysfunction, a condition currently only detectable using different more expensive and time-consuming imaging tests.

Artificial Intelligence Research

There is no inexpensive, noninvasive, painless screening tool for asymptomatic left ventricular dysfunction out there for diagnostic use, researchers said.

The study found that the best existing screening test for asymptomatic left ventricular dysfunction is to measure naturalistic peptide levels (BNP), however results of BNP have been disappointing. The take a look at conjointly needs blood draws.

The system was trained on information from over 600,000 patients, pairing electrocardiogram and echo cardiogram results. The resulting algorithmic program was subsequently tested on an independent data-set comprised of over 50,000 patients. The results were exceptionally positive, with sensitivity, specificity, and accuracy all sitting around 85 p.c. These detection rates were noted as comparable to several alternative popular screening tests, together with breast cancer mammography and prostate cancer antigens.

The study also revealed that in patients without ventricular dysfunction, those with a positive AI screen were at four times the risk of developing future ventricular dysfunction, compared with those with a negative screen. “In other words, the test not only Identified.

asymptomatic diseasehowever also predicted risk of  future diseasepresumablyby  distinctive very early, subtle ekg changes that occur before heart muscle weakness,” said Dr. Friedman.