Published Paper Details:

One of the leading causes of death is lung cancer. The majority of the time, a delayed diagnosis and subsequent therapy result in lung cancer death, claims a study. In a very small percentage of cases where lung cancer is discovered early, it can be treated. Thus, improving the patient's survival depends greatly on the early identification of lung cancer. Based on generative adversarial networks (GANs), this study suggests an approach to lung segmentation accuracy that is more accurate. We may employ GANs and capitalise on their capacity to change images to achieve image segmentation. The ability of generative adversarial networks (GANs) to synthesise globally and locally coherent images with object shapes and textures that are indistinguishable from genuine images is one of the fundamental issues still to be solved. In this study, we offer lung segmentation using U-net, one of the popular deep learning architectures for image segmentation [1]. Therefore, this process is required to remove extra information from lung CT pictures. Based on a limited image collection that includes a few hundred manually segmented lung images, our network accurately segments the data.