A Novel Approach to Chest Disease Detection from Chest X-Ray Images

Abstract

Chest diseases are one of the common diseases in humans, many viral borne diseases also attack the respiratory systems. In such situations, it becomes very important to detect and cure the disease as soon as possible. The chest x-ray is one of the most important sources to detect and identify chest disease. However, detecting the disease can be complicated and may require several medical tests. With the advancement in computer vision technologies, machines can extract information from images. We have trained the computer vision-based models for the task of phenomena disease delectation from the chest x-ray images. In this research paper, we present the novel approach for disease delectation using the ribs extractor framework. The ribs extractor model presented in this research paper was developed using the Conditional generative adversarial network. We have used CNN, densenet, resnet, VGG, and vision, transformer models. We have employed the transfer learning techniques for densenet, resnet, and VGG models. We also present the comparative study of the computer vision models without and with ribs extractor. Finally, we discuss the future scope and suggest ways to improve computer-aided disease detection. We hope that this research helps the research community to better understand medical image centric disease detection.