STEMI Recognition in Electrocardiogram (ECG) by Online AI Bot for Emergency Service and Mobile Medicine

Emergency medical treatment in remote areas often requires immediate medical decision and on-board medical treatments in ambulances before patients could arrive hospitals. For example, a heart attack (Myocardial Infarction) particularly needs prompt treatments to save life. Amid all acute heart diseases, the ST-Elevation Myocardial Infarction (STEMI) is de facto one of the main causes of death. Adequate and immediate treatments for STEMI patients in ambulances before arriving the hospital could largely improve the chance of survival. This paper proposes an online mobile application for real time STEMI ECG pattern recognition. In the cloud AI computing server, a Deep Learning based mechanism is developed to accurately recognize STEMI ECG pattern. On the side of mobile clients, the developed system exploits the salient properties of seamless connections and prompt responses of a wireless mobile network. The main component of our system is an online AI Bot which can help paramedics to recognize STEMI and then to give patients appropriate treatment, such as giving therapeutic drugs and carrying them to the emergency medical center. Collaborating with a teaching hospital, the proposed system is under a testing run and the results show that our system has a very high STEMI recognition accuracy and could be a great help for saving lives of those patients encountering STEMI in remote areas far away from the hospital.