Abstract

Abstract: Pneumonia is derived from the Greek word ‘Pneuma’ which means “Breath”. Pneumonia is a serious diseaserelated to the lungs. Due to Its rapid and intense spread, it has become an important issue that is to be taken care of in thefield of medicine. This disease is more prevalent in people who are addicted to smoking. Results in different domains of imageclassification are remarkable in recent advances achieved in deep learning. In making better decisions there is a dire need forcomputer-aided diagnosis to assist clinicians. Concerned with Pneumonia detection, ResNet; the winner of the IMAGENETchallenge has performed quite well on the image recognition task. ResNet is a classic neural network used as a backbone formany computers vision tasks. A ResNet architecture consists of two residual blocks; an identity block and a convolutionalblock. By using more deeper networks and algorithms, the performance of the models can be improvised. Using X-ray images,work can be extended to classify and both Pneumonia and lung cancer. This paper targets the analysis of detection ofPneumonia based on performance using different computer-aided techniques such as ResNet 18, 34,50,101 and152.Keywords: Pneumonia detection, IMAGENET, ResNet, X-ray imaging