Artificial Intelligence, Diagnostic and Prevention Tools to Deal with Covid 19 In Algeria

In this work, we suggest an original architecture for medical self-diagnosis. Two intelligent models have been created to interact with patients and suggest appropriate medical treatments. This architecture enhances the obtained results through the integration of a Chatbot and an image classification system. The originality of our work lies in the design and development of novel diagnostic procedures. Another unique aspect of our study is the introduction of an architecture that combines text and image classification to identify suitable medications. To validate our proposal, we used COVID-19 disease as a case study, enabling us to guide patients with the assistance of an intelligent agent. Our model was trained using Convolutional Neural Networks (CNNs) on a dataset comprising three classes: COVID, normal, and viral pneumonia. We introduced a new stopping criterion based on accuracy, effectively reducing learning time and preventing confusion. In the latter part of this paper, we examined the consumption patterns of anticoagulants, antibiotics, anti-inflammatories, and analgesics over two years: pre-COVID-19 (2018-2019) and through COVID-19 (2020-2021). Our dataset was compiled from data collected at Mohamed Boudiaf Barika Hospital in Algeria. Based on this study, we proposed a predictive system for drug consumption trends.