A special computer algorithm based on artificial intelligence (AI) created by Mount Sinai researchers can learn how to recognize subtle changes in an electrocardiogram (also called an electrocardiogram or electrocardiogram) to predict whether a patient is experiencing heart failure.

"We show that deep learning algorithms can identify blood pumping problems on both sides of the heart from ECG waveform data," said Dr. Benjamin S. Glicksberg, assistant professor of genetics and genomics science and member of the Hasso Plattner Institute for Digital Research, of Mount Sinai’s health, The senior author of the study was published in the Journal of the American College of Cardiology: Cardiovascular Imaging. "Usually, diagnosing these types of heart diseases requires expensive and time-consuming procedures. We hope that the algorithm can diagnose heart failure faster."

The research was led by Akhil Vaid, MD, Girish N. Nadkarni, a postdoctoral scholar in the Glicksberg laboratory, MD, Master of Public Health, CPH, Associate Professor of Medicine at Icahn School of Medicine at Mount Sinai, and Director of the Department of Medicine Data Driven and Digital Medicine (D3M) , And the senior author of the study.

Heart failure or congestive heart failure, which affects approximately 6.2 million Americans, occurs when the heart pumps less blood than the body normally needs. For many years, doctors have relied heavily on an imaging technique called echocardiography to assess whether a patient may be experiencing heart failure. Although useful, echocardiography can be a labor-intensive procedure and is only available in certain hospitals.

However, recent breakthroughs in artificial intelligence have shown that the electrocardiogram-a widely used electronic recording device-may be a fast and easily available alternative in these situations. For example, many studies have shown that "deep learning" algorithms can detect weaknesses in the left ventricle of the heart and push fresh oxygenated blood to other parts of the body. In this study, the researchers described the development of an algorithm that can evaluate not only the strength of the left ventricle, but also the strength of the right ventricle. The algorithm can absorb deoxygenated blood from the body and pump it to the lungs. Department.

"Although it is attractive, it has been challenging for doctors to use ECG to diagnose heart failure traditionally. This is partly because there are no established diagnostic criteria for these assessments, and some changes in ECG readings are too subtle for the human eye." Dr. Nadkarni said. "This research represents an exciting step forward in discovering the information hidden in ECG data, which can use relatively simple and widely available tests to obtain better screening and treatment paradigms."

Usually, the electrocardiogram includes a two-step process. The wires are attached to different parts of the patient's chest, and within a few minutes, a specially designed portable machine will print out a series of wavy lines or waveforms that represent the electrical activity of the heart. These machines can be found in most hospitals and ambulances in the United States and can be operated with minimal training.

In this study, the researchers programmed the computer to read the patient's electrocardiogram and data extracted from written reports, which summarized the results of the corresponding echocardiograms collected from the same patient. In this case, the written report serves as the computer's standard data set, which can be compared with ECG data and learn how to spot a weaker heart.

Natural language processing programs help computers extract data from written reports. At the same time, a special neural network that can discover image patterns is combined to help the algorithm learn to recognize pumping intensity.

"We hope to advance the most advanced technology by developing artificial intelligence that can understand the entire heart easily and cheaply," said Dr. Vaid.

The computer then read more than 700,000 ECG and echocardiogram reports obtained from 150,000 Mount Sinai Health System patients from 2003 to 2020. The data from the four hospitals were used to train the computer, and the data from the fifth hospital was used to test the execution of the algorithm in different experimental environments.

"One potential advantage of this research is that it involves one of the largest ECG collections from one of the most diverse patient populations in the world," said Dr. Nadkarni.

The initial results show that the algorithm can effectively predict which patients have a healthy or very weak left ventricle. The intensity here is defined by the left ventricular ejection fraction, which is an estimate of how much fluid is pumped by the ventricle during each heartbeat observed on the echocardiogram. The ejection fraction of a healthy heart is 50% or higher, while the ejection fraction of a weak heart is equal to or lower than 40%.

The algorithm is 94% accurate in predicting which patients have healthy ejection fractions, and 87% in predicting patients with ejection fractions below 40%.

However, the algorithm is not as effective in predicting which patients will have a slightly weakened heart. In this case, the program is 73% accurate in predicting patients with ejection fractions between 40% and 50%.

Further results show that the algorithm has also learned to detect the weakness of the right valve from the electrocardiogram. In this case, the weakness is defined by more descriptive terms extracted from the echocardiogram report. Here, the algorithm is 84% ​​accurate in predicting which patients have weaker right valves.

"Our results indicate that the algorithm may eventually help doctors correctly diagnose failure on both sides of the heart," said Dr. Vaid.

Finally, additional analysis shows that the algorithm can effectively detect cardiac weakness in all patients, regardless of race and gender.

"Our results indicate that the algorithm may be a useful tool to help clinical practitioners fight heart failure in various patients," Dr. Glicksburg added. "We are carefully designing forward-looking trials to test their effectiveness in a more realistic environment." 

This research was supported by the National Institutes of Health (TR001433).

Vaid, A. et al. Using deep learning algorithms to simultaneously identify left and right ventricular dysfunction from ECG, Journal of the American College of Cardiology: Cardiovascular Imaging, October 13, 2021, DOI: 10.1016/j.jcmg.2021.08.004.

About Mount Sinai Health System

The Mount Sinai Medical System is the largest academic medical system in New York City, including eight hospitals, a leading medical school, and a large network of outpatient clinics throughout the New York area. Mount Sinai promotes the development of medicine and health through unparalleled education and translational research and discovery to provide the safest, highest quality, most accessible and fair care, and the most valuable care in any health system in the country. The health system includes approximately 7,300 junior and specialist physicians; 13 joint-venture outpatient surgery centers; more than 415 outpatient treatments in five boroughs of New York City, Westchester, Long Island, and Florida; and more than 30 affiliated community health center. Mount Sinai Hospital ranks among the top 20 hospitals in the United States on the "Honor Roll" of U.S. News and World Report, and ranks first in the country: No. 1 in Geriatrics, Cardiology/Cardiac Surgery, Diabetes/Endocrinology, Stomach Top 20 in Enterology/Gastrointestinal Surgery, Neurology/Neurosurgery, Orthopedics, Pulmonology/Pulmonary Surgery, Rehabilitation and Urology. The Ophthalmology Department of Mount Sinai New York Eye and Ear Hospital ranked 12th. In the "Best Children's Hospital" by U.S. News and World Report, the Kravis Children's Hospital of Mount Sinai ranks among the best in four countries out of ten pediatric specialties. Icahn School of Medicine is one of the three medical schools that have won awards through multiple indicators: ranked in the top 20 in the "Best Medical Schools" by U.S. News and World Report, and tied with U.S. News and World Report's "Honour Roll" hospitals , Supported by the 14th National Institutes of Health in the country. Newsweek’s "Best Smart Hospital in the World" ranked Mount Sinai Hospital as the number one in New York and the top five in the world, and Mount Sinai Morningside Hospital in the top 20 in the world.

For more information, please visit https://www.mountsinai.org or find Mount Sinai on Facebook, Twitter and YouTube.

If you are a member of the media and before the deadline, please call the press office or call the press officer on call.

If you are a member of the media and before the deadline, please call the press office or call the press officer on call.