Improving Prognosis for Pneumonia using Deep Learning Techniques
Abstract
Pneumonia, a lung infection, when not treated correctly, maybe a life-threatening illness. New technologies, in addition to advances in medical science, play a crucial role in many attempts to diagnose the disease. In the area of image recognition, several computer vision researchers are contributing and they have suggested a wide variety of solutions. Thanks to its extremely robust non-linear architecture, deep convolution networks are among all of these approaches achieving outstanding efficiency in object detection and classification. The current research proposes a deep convolution neural network architecture with a view to enhancing the accuracy of pneumonia detection from an x-ray of the chest. In the proposed approach the x-ray images are classified using a custom model and are compared with popular CNN-based deep network VGG-16. A comprehensive analysis of how these algorithms operate and how they are compared is given by us. The dataset is composed of 5863 two-class (normal and pneumonia) images.