Extraction and assessment of COVID19 infection in lung CT images using VGG-UNet

The infectious disease in humans is gradually rising for various reasons, and COVID19 is one of the recently discovered diseases caused by SARS-CoV-2. From early 2020, the infection due to COVID19 has gradually increased, and still, its infection exists. COVID19 will cause severe infection in the respiratory tract, and early detection and treatment are essential. The harshness of the infection needs to be examined before implementing the treatment. This research aims to build up and implement a suitable procedure to extract and assess the infected section in lung CT slices. This work extracts the infected section using the pre-trained VGG-UNet scheme. The separated section is validated against the ground-truth (GT) image, and the necessary presentation standards are calculated. The performance of the VGG-UNet is then compared and verified with the UNet and UNet+ schemes. The investigational product of this study authenticate that the effect reached with the proposed study confirms that the VGG-UNet provides better Jaccard, Dice and accuracy compared to UNet and UNet+.