Multi-Purpose Chest X-Ray Analytics System Using Deep Learning Techniques

There has been rapid and tremendous progress in the past few years in the field of deep learning, mainly due to the availability of computational power and the vast amount of data. Deep neural networks have resulted in state-of-the-art performances in classification, detection, and segmentation in comparison to previous shallow methodologies built upon hand-crafted image features. This progress has yielded a rise in performance of medical imaging analysis as well. Chest X-ray imaging is one of the most accessible medical imaging technique for diagnosis of multiple diseases. With the availability of ChestX-ray14, which is a massive dataset of chest X-ray images, we can use deep learning techniques for various tasks like classification, detection, segmentation, etc. This thesis explores deep learning techniques such as convolutional neural networks and generative adversarial networks for applications of content-based image retrieval, image generation for different pathologies, and recognition of pathologies in chest X-rays.