Analysis of Machine Learning Algorithms for COVID Detection Using Deep Learning

Abstract

The COVID-19 worldwide epidemic has disrupted our lives in countless ways. An ongoing pandemic, COVID-19 (coronavirus) is brought on by coronavirus 2 (severe acute respiratory disease) (SARS-CoV-2). In more specific terms, the limits of the healthcare systems were reached. Artificial intelligence has made great strides recently, making it possible to create sophisticated applications that satisfy requirements for clinical accuracy. It is crucial to develop and implement an automatic detection system that can be applied on a commercial scale to provide for an alternate diagnosis method for COVID-19 detection since medical institutions only have a limited few of COVID-19 test kits. A chest X-ray is the initial imaging technique utilized to identify COVID-19 illness. Convolutional neural network-based deep learning models that were specially created and previously trained were utilized to identify pneumonia brought on by COVID-19 respiratory issues. Data from a publicly accessible dataset was utilized. Large-scale annotated picture dataset availability has successfully made convolutional neural networks for image analysis and classification. We proposed a deep convolutional neural network in this study that has been trained on two open-access datasets, namely Normal and COVID. The outcomes demonstrate a high detection precision of 95% for the combined dataset, and most models handled new data with only a little decrease in accuracy.