Deep Learning Approach for Automatic Classification of X-Ray Images using Convolutional Neural Network

Medical Image diagnosis has increased drastically in recent years. It may be a relatively simple task to read and diagnose chest X-ray images but the complexity of the images sometimes lead to improper diagnosis. After the success of deep learning in many industries, it also has proven results in the form of better accuracies for medical imaging. In this paper, the authors used Xception model which is a pre-trained model of the state-of-the-art deep learning image classifiers. Our deep learning approach has helped to build highly accurate prediction model for the identification of different diseases from chest X-rays. The classification accuracy will benefit the patients who do not have access to radiologists to read their chest xrays. The experimentation has been done using the Keras library built under Tensorflow backend.