An optimized CNN based automated COVID-19 lung infection identification technique from C.T. images

The Novel Coronavirus, commonly known as COVID-19, is a highly contagious disease that spreads all over the globe has brought great suffering. The symptoms have made all ages suffer a lot. The diagnostics of COVID-19 is an excellent challenge for the medical field as the mutated form of the virus gives out its symptoms in different forms. The main diagnostics to be involved in this infection of COVID-19 is the Lung operation. Especially the automatic detection of the lungs' infection using chest X-rays provides the comprehensive possibility for healthcare professionals to develop hospital procedures to handle this COVID-19. Computed tomography scans are used to diagnose the lungs' infection caused by the Coronavirus, whereas the C.T. scans break the infected region from lung lesions. It is imperative to measure the disease progression, which is challenging to track down and treat accurately. Currently, Segmenting the contaminated regions from the C.T. slices can create loads of problems, which involves more alteration in contamination nature and minor power disagreement in the center of the infected tissues and the typical material. This chapter aims to segment the infection in the lungs using SqueezeNet as the Convolutional Neural Network (CNN) to recognize the contaminated regions automatically. This approach may be useful to help in the accuracy of the C.T more accurately. It has been ensured based on Dice scores, sensitivity, and high precision, specificity. The results achieved with the proposed model are low for the former and directly proportional to the latter compared with existing methods.