Prediction of Covid-19 Infected Chest X-Rays Using Machine Learning Techniques

Abstract

As the number of cases with COVID-19 continues to climb, better medical screening and clinical care of the condition
are urgently needed. Typical signs of a chest cold include a sore throat, cough, and a high temperature. Patients with pneumonia also have these symptoms. Because of this, detecting COVID-19 is much more difficult. Recognizing COVID-19 in a set of Chest-X-Ray (CXR) pictures that also contains pneumonia patients is laborious and prone to human mistake. Radiography of the chest for COVID-19 cases and others with comparable symptoms may be used as a first-line triage method. While this is true, radiologists still have a hard time distinguishing between COVID-19CXR pneumonia and other types of pneumonia due to the similarities in their appearance. This work is an effort to construct a machine learning model that is beneficial in categorizing CXR pictures into three classes indicating normal, COVID-9, and pneumonia based on the premise that such classifiers can consistently identify COVID-19 CXR images from other kinds of pneumonia. Feature The methods of extraction, dimensionality reduction, and machine learning are all used.