Contribution to chest Radiograph Pathology categorization

Abstract

Classification of chest diseases is one of the most interesting research topics in recent years because it requires rapid, high-accuracy diagnosis. Although chest radiography has several advantages in diagnosing Radiologists have always been specialists in the field of diseases, but the process of understanding the image is a big problem for doctors and Radiology due to diagnostic errors made by experts in the field, and for this reason it is still the diagnostic process Somewhat confusing and difficult. This encouraged the use of modern artificial intelligence techniques such as deep learning to diagnose chest diseases and classification from the introduction of medical images and x-ray images, for this purpose began Deep learning scientists are involved in building systems based on deep learning, more precisely devolutionary neural networks. In this senior project, we have created three deep learning models based on analytical neural networks to classify some chest diseases from X-ray images. The first model is a binary classifier Related to Corona virus cases (patient/normal) achieved an accuracy rate of 99%, the second model related to inflammation Pneumonia rate of 96%, and the third model on tuberculosis with a user interface score of 98%.