Lung’s Segmentation Using Context-Aware Regressive Conditional GAN

After declaring COVID-19 pneumonia as a pandemic, researchers promptly advanced to seek solutions for patients fighting this fatal disease. Computed tomography (CT) scans offer valuable insight into how infection affects the lungs. Analysis of CT is very significant, especially when physicians are striving quick solutions. This study successfully segmented lung due and provided physician with quantitative analysis condition. lesions often occur near over parenchyma walls, which denser exhibit lower contrast than tissues outside parenchyma. We applied Adoptive Wallis Gaussian filter alternatively regulate outlining lungs proposed context-aware conditional generative adversarial network (CGAN) gradient penalty spectral normalization automatic segmentation lesion segmentation. The CGAN implements higher-order statistics compared traditional deep-learning models. produced promising results Similarly, has shown outstanding an accuracy 99.91%, DSC 92.91%, AJC 92.91%. Moreover, we achieved 99.87%, 96.77%, 95.59% Additionally, suggested attained sensitivity 100%, 81.02%, 76.45%, 99.01%, respectively, critical, severe, moderate, mild severity levels. model outperformed state-of-the-art techniques detection cases.