Zhao Yifan, Xiong Jing, Hou Yang, Zhu Mengyun, Lu Yuyan, Xu Yuanxi, Teliewubai Jiadela, Liu Weijing, Xu Xiao, Li Xin, Liu Zheng, Peng Wenhui, Zhao Xianxian, Zhang Yi, Xu Yawei
Department of Cardiology, Shanghai Tenth People's Hospital, Tongji University School of Medicine, Shanghai, China.
Shanghai MobileVision Technology Co., Ltd, China.
Int J Cardiol. 2020 Oct 15;317:223-230. doi: 10.1016/j.ijcard.2020.04.089. Epub 2020 May 3.
Patient delay is a worldwide unsolved problem in ST-segment elevated myocardial infarction (STEMI). An accurate warning system based on electrocardiogram (ECG) may be a solution for this problem, and artificial intelligence (AI) may offer a path to improve its accuracy and efficiency. In the present study, an AI-based STEMI autodiagnosis algorithm was developed using a dataset of 667 STEMI ECGs and 7571 control ECGs. The algorithm for detecting STEMI proposed in the present study achieved an area under the receiver operating curve (AUC) of 0.9954 (95% CI, 0.9885 to 1) with sensitivity (recall), specificity, accuracy, precision and F1 scores of 96.75%, 99.20%, 99.01%, 90.86% and 0.9372 respectively, in the external evaluation. In a comparative test with cardiologists, the algorithm had an AUC of 0.9740 (95% CI, 0.9419 to 1), and its sensitivity (recall), specificity, accuracy, precision, and F1 score were 90%, 98% and 94%, 97.82% and 0.9375 respectively, while the medical doctors had sensitivity (recall), specificity, accuracy, precision and F1 score of 71.73%, 89.33%, 80.53%, 87.05% and 0.8817 respectively. This study developed an AI-based, cardiologist-level algorithm for identifying STEMI.
患者延误是ST段抬高型心肌梗死（STEMI）领域一个全球范围内尚未解决的问题。基于心电图（ECG）的精确预警系统可能是解决该问题的一个办法，而人工智能（AI）或许能为提高其准确性和效率提供一条途径。在本研究中，利用一个包含667份STEMI心电图和7571份对照心电图的数据集，开发了一种基于AI的STEMI自动诊断算法。本研究中提出的检测STEMI的算法在外部评估中，受试者工作特征曲线下面积（AUC）达到0.9954（95%CI，0.9885至1），灵敏度（召回率）、特异性、准确率、精确率和F1分数分别为96.75%、99.20%、99.01%、90.86%和0.9372。在与心脏病专家的对比测试中，该算法的AUC为0.9740（95%CI，0.9419至1），其灵敏度（召回率）、特异性、准确率、精确率和F1分数分别为90%、98%、94%、97.82%和0.9375，而医生的灵敏度（召回率）、特异性、准确率、精确率和F1分数分别为71.73%、89.33%、80.53%、87.05%和0.8817。本研究开发了一种基于AI的、达到心脏病专家水平的识别STEMI的算法。
