A Detailed Study on AI-Based Diagnosis of Novel Coronavirus from Radiograph Images

Abstract

Novel coronavirus 2019 (COVID-2019) initially started at Wuhan, China, and spread all over the world and was announced as a pandemic by the World Health Organization in March 2020. This virus causes pneumonia in human beings and changes the respiratory pattern (different from common cold and flu). Compared to the reverse-transcription polymerase chain reaction, chest X-ray and computed tomography imaging may be reliable and quick to diagnose the COVID-19 patients in the epidemic regions. Above-mentioned imaging modalities along with machine learning techniques can be helpful for accurate diagnosis of the disease. Further, machine learning can improve throughput by accurate figuration of contagious X-ray and tomography images for early detection and its subsequent impact on the respiratory system. In this chapter, the details of medical imaging and artificial intelligence-based automatic classification techniques for COVID-19 have been discussed and analyzed. Here, a 24-layer convolutional neural network has been proposed for the binary (COVID vs. non-COVID) and multi-class (COVID vs. non-COVID vs. pneumonia) classification from X-ray and tomography images. In two classes, it attains an accuracy of 71.81% and 99.68% on tomography and X-ray images, respectively. Precision is 100% for the X-ray dataset, with an overall accuracy of 96.4% for multi-class classification in X-ray images.