Quick diagnosis of acute myocardial infarction with electrocardiogram AI...Accuracy 87.8%
Mar 13, 2025
|
Professor Kim Kyu-seok of the Department of Emergency Medicine at Bundang Cha Hospital of Cha Medical University (corresponding author) published the results of a joint study with Dr. Lee Min-sung of Medical AI (first author, emergency medicine specialist), Professor Shin Tae-gun of the Department of Emergency Medicine at Samsung Medical Center (first author), Professor Lee Young-joo of the Department of Emergency Medicine at Soonchunhyang University Seoul Hospital (first author), and CEO Kwon Joon-myung of Medical AI (corresponding author, emergency medicine specialist) in the official journal of the European Heart Association 'European Heart Journal (38 points of IF), the best journal in the world's cardiovascular system.
Acute myocardial infarction is one of the four major emergency diseases, which is caused by sudden blockage of cardiovascular vessels that supply oxygen and nutrients to the heart, and has a mortality rate of 5-10%. Early diagnosis is very important because the mortality rate increases sharply as diagnosis and treatment are delayed.
In response, Professor Kim Kyu-seok's team measured the accuracy of diagnosis of acute myocardial infarction through the artificial intelligence electrocardiogram analysis program developed by MedicalAI for 8,500 patients who visited 18 university hospitals in Korea from March 2022 to October 2023 with chest pain and compared it with Physician AMI, HEART, and GRACE 2.0. As a result, the accuracy of the AI electrocardiogram was 87.8%, higher than that of doctors (84.6%).
It is considered an innovative diagnostic method in that it can determine acute myocardial infarction only by electrocardiogram in the early stages. The initial electrocardiogram can accurately determine whether there is acute myocardial infarction in 21.4% of patients, and when artificial intelligence electrocardiogram analysis and initial blood test results were collected, 51.8% of patients were diagnosed with acute myocardial infarction early.
Professor Shin Tae-gun of Samsung Medical Center and Samsung Medical Center, a co-author, said, `It is very surprising that AI electrocardiogram analysis has a higher diagnostic ability than the overall judgment obtained by face-to-face with patients such as their medical history and physical examination.'
Professor Kim Kyu-seok, co-corresponding author of Bundang Cha Hospital, said, `Although the development of programs using artificial intelligence is currently being actively carried out in medical care, it is often necessary to evaluate the accuracy in the actual medical environment because it is an analysis result conducted with retrospective data. In this study, I think it is a very meaningful result because it was obtained by driving the developed program directly in medical reality," he said.
Meanwhile, this study was conducted by Medical AI and the Korea Health Industry Promotion Agency's artificial intelligence task support.
|
This article was translated by Naver AI translator.
