Generative Adversarial Network (GANs) based training set enhancement for Stomach Adenocarcinoma Computed Tomography (CT) scan

The new era of Deep learning has found many widespread uses in diverse fields such as Science, Fashion, Game, Medical, Health etc. which has gained a huge attention for the researchers. Recently, Generative Adversarial Network (GAN) has crucial contribution in the field of medical image analysis, along with different variants of GAN it enhances the capability to resolve the challenging problems in medical field which leads to the betterment of healthcare technologies. Furthermore, GAN has proven to be useful, to synthesize images that can resolve the scarcity of real training data especially in medical and healthcare field of research. In this paper, we present an attempt to improve dataset for CT scan images of stomach cancer using GAN approaches to various fields of medical image analysis. The paper also provides a statistical comparison of generated images which are found to be of high quality.