ABSTRACT
The COVID-19 outbreak has thrown the entire world into an unanticipated unpleasant scenario, bringing the lives of people all over the world to a pandemic level and claiming thousands of lives. According to WHO, COVID-19 has spread to 220 nations and territories, with the number of infected cases and deaths reaching 167 million and 3 million (as of May 25, 2021) (WHO dashboard, https://covid19.who.int [1]) and has a serious impact on the public health system. The key hurdles in containing the present COVID-19 outbreak are early detection and diagnosis. As a result, it is critical to screen COVID-19-affected individuals as soon as possible. Otherwise, it will spread quickly. In this case, screening can determine whether or not a patient has COVID-19 pneumonia. One of the most effective methods for reaching this goal is through chest X-ray diagnosis. It is the one that is most easily detected. This study provided a deep learning model of a convolutional neural network solution for detecting COVID-19 pneumonia patients using chest X-ray pictures. We used a publicly accessible chest X-ray dataset from Kaggle (Dataset link, https://www.kaggle.com/tawsifurrahman/covid19-radiography-database [2]) to train the model, which included 10,006 photos with COVID-19 pneumonia and normal images separated into train, test, and validation sets. This proposed model has a classification precision of 96% on the test set and 97% on the validation set, which is rather good for classifying COVID-19 pneumonia and normal patients. Along with this, we add the functionality of suggesting medicine based on symptoms. We build this suggesting model using machine learning model with accuracy 79% of 120 rows of dataset developed by our own to show it is possible to suggest medicine based on symptoms. We have developed an application using the flask framework. This application may be used on any computer by any medical professional to detect COVID positive and negative patients automatically using chest X-ray images in a matter of seconds and recommends some medicine that currently threatens the COVID-19. This application can reduce the number of false positives and false negatives in the detection of COVID-19. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.