Automatic Detection of COVID 19 Infection Using Deep Learning Models from X-Ray Images. Issue 1 (March 2021)

• Record Type: Journal Article
• Authors: Yadav, Anju; Verma, Vivek Kumar; Pal, Vipin; Singh, Saumya
• Is Part Of: IOP conference series. Volume 1099:Issue 1(2021)
• Publication Date: 2021-03
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/1099/1/012050
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: In worldwide millions of people are infected due to COVID-19 virus and even that is increasing exponentially. The medical researchers are using RT-PCR test to distinguish between COVID (+) and COVID (-). The growth of virus is exponential that's lead to an outbreak, and to break this we have to conduct test as many as possible. RT PCR test is costly and not available in remote areas this is big challenge that developing countries is facing. The second big challenge whole world facing is the availability of doctors and radiologists. So, we need an automated system based on AI and Machine learning (ML) for detection of COVID (+) cases. In literature AI, ML and Deep Learning (DL) played an important role in designing automated system in medical field by achieving good accuracy. In this paper, we have applied DL models ResNet 34 and ResNet 50 on X-ray images, achieved accuracy 96.4% with ResNet 34. And compared their result on the basis of various quantitative analysis parameter MCC, F1 Score, Precision, Recall and others to solve both challenges.