A detailed study on the covid-19 outbreak in india and a deep learningbased approach to detect lung infection using CT images

On December 31, 2019, World Health Organization (WHO), China country office, was informed about new cases of pneumonia with unknown factors detected in Wuhan City hospital, Hubei province of China. Resources say that the outbreak of this epidemic novel coronavirus (nCoV) was said to have spread from an animal source to humans from Wuhan's Huanan seafood market, China. COVID-19 detected in China is closely related to the severe acute respiratory syndrome (SARS-CoV-1) virus and it is also named SARS-CoV-2 virus. While animals are believed to be the source of spread, the virus is now spread commonly from person to person and claimed thousands of lives across the globe. The onset of this symptom raged between December 8, 2019, to January 12, 2020, and within this period, a total of 198 cases were identified with nCoV in China. Out of these 198 cases, 136 cases were identified in Wuhan, 2 in Beijing, 1 in Guangdong, and 2 in Shanghai. During this period, nine exported cases were reported from different countries, such as one in Japan, two in Thailand, and one in South Korea. Of these reported cases, 43 were severe illnesses, 9 were said to be critical, and the remaining were stable under medical observation. On January 30, 2020, the first confirmed case of COVID-19 was diagnosed in a laboratory in Kerala, India, from an infected person's sample. This patient was a student returning from Wuhan. Day by day, the number of active cases and death rates was mounting tremendously creating a pandemic outbreak of the SARS-COVID-2 virus in most countries. This sudden outbreak of COVID-19 has caught many countries unprepared for treating the illness and they are struggling to find an antidote for the virulent novel coronavirus. Since COVID-19 is a disease with no immediate cure the first resort was early detection of the virus and isolation of the patients. RT-PCR kits were used as the primary method of diagnosing COVID-19. It has been observed that these tests produced results indicating the patient doesn't have a virus when they do. The rates of false negatives have ranged from 2% to 37% for RT-PCR kits whereas it is around 50% for the antigen tests. CT scans were used for a more accurate result reducing the inefficiency of the initial tests. The prices and need for chest X-rays and HRCTs have exponentially shot up due to the high demand for these tests. A CT or computerized tomography scan makes use of computers and rotational X-ray machines for creating cross-sectional images of the human body. The application of deep learning in the field of COVID-19 radiologic image processing reduces false-positive and negative errors in the detection and diagnosis of this disease and offers a unique opportunity to provide fast, cheap, and safe diagnostic services to patients. In our chapter, we aim to build a convolutional neural network to detect COVID in a CT scan. Extensive training of the network using our models has achieved a training accuracy of 96.04% and a testing accuracy of 95.17%