AUTHOR=Wu Lin , Huang Guifang , Yu Xianguan , Ye Minzhong , Liu Lu , Ling Yesheng , Liu Xiangyu , Liu Dinghui , Zhou Bin , Liu Yong , Zheng Jianrui , Liang Suzhen , Pu Rui , He Xuemin , Chen Yanming , Han Lanqing , Qian Xiaoxian 

TITLE=Deep Learning Networks Accurately Detect ST-Segment Elevation Myocardial Infarction and Culprit Vessel

JOURNAL=Frontiers in Cardiovascular Medicine

VOLUME=Volume 9 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/cardiovascular-medicine/articles/10.3389/fcvm.2022.797207

DOI=10.3389/fcvm.2022.797207

ISSN=2297-055X

ABSTRACT=Early diagnosis of acute ST-segment elevation myocardial infarction (STEMI) and determination of the culprit vessel is associated with a better outcome. We developed three deep learning (DL) models for detecting STEMI and culprit vessel based on 12-lead electrocardiography (ECG), and compared them with experienced doctors including cardiologists, emergency physicians and internists. After screening the coronary angiography results, 883 cases (506 control and 377 STEMI) from internal and external datasets were enrolled for testing DL models. CNN-LSTM (Convolutional Neural Network-Long Short-Term Memory) (AUC:0.99) performed better than CNN, LSTM and doctors in detecting STEMI. DL models (AUC:0.96) performed similarly to experienced cardiologists and emergency physicians in discriminating the left anterior descending artery. In regarding to distinguish RCA from LCX, DL models were comparable to doctors (AUC:0.81). In summary, we developed ECG-based DL diagnosis systems to detect STEMI and predict culprit vessel occlusion, thus enhancing the accuracy and effectiveness of STEMI diagnosis.