Published Paper Details:

After its first reported occurrence on 31st December 2019 and getting its name on Feb 11, 2020, Covid-19 was declared a Pandemic on March 11, 2020, by the World Health Organization (WHO). The highly contagious nature of the disease along with its incompletely understood transmission process and the absence of a definitive cure made it extremely difficult for World Governments and their Health care providers to effectively contain the infection. In extreme cases, Covid-19 has been observed to result in severe respiratory discomfort, multiple organ failure, and eventually death. Covid-19 based Pneumonia was observed in a small percentage (2% to 8%) of the infected patients which was found to be more fatal compared to regular pneumonia. RT-PCR is the current recommended standard for the detection of Covid-19, but RT-PCR has been proven to be oblivious to early stages of Covid-19 Pneumonia and is associated with low sensitivity, reportedly as low as 30% to 70%. Chest CT and X-rays exhibit superior sensitivity over RT-PCR but cannot differentiate between Covid-19 and Non Covid-19 pneumonia. Therefore, this paper proposes the use of Artificial Intelligence (AI) to assist in the detection and classification of Covid-19 pneumonia from Non Covid-19 pneumonia and No Covid using X-Ray images. A 2D Deep Learning CNN architecture is proposed for solving Binary classification (Covid-19 vs Normal) and Multiclass Classification (Covid-19 pneumonia vs Non Covid-19 pneumonia vs Normal) problem of a custom created diverse X-ray dataset.