Asistente virtual de telemedicina con dispositivo electrónico de medición de triaje para el diagnóstico médico de infecciones respiratorias utilizando inteligencia artificial

Abstract

The symptoms of COVID-19 and other respiratory infections are very similar, complicating the diagnosis of these diseases since they have common symptoms. For this reason, a virtual telemedicine assistant was implemented with electronic triage measuring device for medical diagnosis of respiratory infections using artificial intelligence, which provides personalized and timely care to each patient. The project was developed in collaboration with a health professional who provided guidelines for the measurement of vital signs, as well as the process of diagnosis of a patient. In the case of the electronic device, it has two ESP8266 microcontrollers that are responsible for measuring the heart rate, oxygen saturation in the blood and body temperature. This data is displayed by an Oled screen and sent to the Google Cloud server using the MQTT protocol. Inside the server is hosted the web application consisting of a login screen, which will verify if the user is registered. Once the user logs in, they will access to an interface where the measurements of the vital signs are shown in real time. After the review of the vital signs the user can enter to the screen of the virtual assistant (chatbot) that has natural language processing to simulate the consultation with a real doctor and interpret the symptoms entered by the user to be stored in the MySQL database. At the end of entering the symptomatology, the artificial intelligence algorithm starts with a precision of 0.91, which is responsible for determining whether the patient has covid19, common cold or allergic rhinitis, these three classes were suggested by the doctor. Finally, the user will move to an interface where the data of all the queries made to the chatbot is found. It is worth mentioning that the web application was developed using the Flask framework with the tools of HTML, CSS, Javascript, SQL and Bootstrap 5, to make the application responsive, dynamic and robust, since it has security for both the user and the server