A novel method for detecting ST segment elevation myocardial infarction on a 12-lead electrocardiogram with a three-dimensional display

Abstract

ObjectiveTwo main factors affect the accurate diagnosis of myocardial infarction (MI). The first is achieving accuracy in extracting electrocardiogram (ECG) time domain features and the second is proper analysis of the entire 12-lead ECG. ECG contains measurement data that reflect cardiac activity. However, accurate extraction of ECG features to identify cardiac abnormalities is affected by noise that can degrade the robustness of feature extraction. Each lead in the 12-lead ECG provides different information that can help diagnose MI. Although analyzing the entire 12-lead ECG improves accuracy, it is very time-consuming. This study aimed to propose a three-dimensional (3D) ECG method for detecting ST-segment elevation MI (STEMI) and to confirm the efficacy of the 3D ECG method. MethodsThe leads of the existing 12-lead ECG were categorized into limb lead and chest lead groups and reconstructed them in a 3D format using a 3D display method. We extracted new features from the 3D ECG plane and evaluated their ability to detect STEMI based on new features using the K-nearest neighbor classifier. ResultsThe 3D ECG detected STEMI with 96.37% accuracy. ConclusionThe proposed 3D ECG method can be easily used to read the entire 12-lead ECG even with the naked eye. SignificanceThis method can detect STEMI by simply using the 3D ECG features, thus eliminating the need to extract features such as the ST segment via complicated calculation processes.