Diagnostic imaging for COVID-19 using X-ray

ABSTRACT

The pandemic (COVID-19), which emerged in late December 2019 in Wuhan, China, is still continuing to ravage every country in the world. As a result of the outbreak, the world has experienced tough times. To date, only 24% of the world's population has been vaccinated yet, and over 18 million cases are still active. In many places, the hospital is running short of beds and oxygen cylinders. With many undetected cases and casual approaches by people, this third wave is predicted to be a big hit. Conventional methods of diagnosis, such as antigen analysis, serological tests, and polymerase chain reactions, although widely used, are time-consuming. The use of Deep Learning (DL) and a convolutional neural network (CNN) to examine chest CT (computerized tomography) or chest x-ray images has been shown to be a promising technique for early diagnosis of COVID-19. In this study, a multi-level analysis method is proposed to detect COVID-19 from chest radiographs of a human. Through the model, a fast and accurate diagnosis of coronavirus can be made at the reach of fingertips. Based on patient clinical data, an ANN is used to calculate the likelihood of the patient becoming infected with COVID-19. The proposed model makes use of 8 layers of Convolutional Neural Network which are trained on the dataset (80-10-10 train-validate-test split) which gives an accuracy of 97% on the training data and 98.7% on the validation data. © 2021 IEEE.