In the case of an epidemic, the transfer of patients to emergency departments and their physiological parameters remains the major challenge in this critical situation. Rapid and effective diagnosis can save the lives of these patients. This goal is achieved by using deep learning in classifying information into emergencies to send it first. To do this, we propose in this article, an intelligent emergency system for the rapid transmission of information from emergency patients to the hospital that has the best health care for these patients using a deep learning approach. First, the fusion method combines data collected and filtered from patients' electronic medical records with data captured by the wireless medical sensor network during the transport phase. Next, the method of selecting the parameters is used as inputs to the learning model. The information collected and learning outputs, such as emergency alerts are sent via Wi-Fi and 5G equipment in our intelligent system. Accuracy of our proposed contribution reaches 98% with 1.53s run time. This finding shows that our system is effective and is used in cases of epidemics as COVID-19.