A Semi-supervised Pipeline for Detection of Anomalies in Medical Images

With the breakthroughs in the field of deep learning and computer vision problems, manyworkflows have been revolutionized, including diagnostic medical imaging. In this work we developed a Generative Adversarial Network - based pipeline for augmentation of data and detection of anomalies in the form of Viral Pneumonia and COVID-19 in chest X-ray images. By eliminating the need for more labelled data for training our model, we offered a pipeline that reduced the strain on radiologists when curating data for deep learning models. We demonstrated that our pipeline improves results compared to the previous state-of-the-art generative model for detection of anomalies in the images.M.Sc