Deep Neural Network Approaches for Pneumonia Identification in Chest X-Rays

Pneumonia is a type of infection that affects the lungs and in some cases may be life-threatening. The images of chest x-ray can give useful [1]information for diagnosis and cure plans of pneumonia. So, training and building a model that can label a given chest x-ray image can prove to be of great use in the medical field. The current research is focused at enhancing the accuracy of the pneumonia analysis using CNN. We will also test other various architectures such as RESnet, Googlenet, VGGnet, and Alexnet. We do so by examining their effectiveness in feature extraction and image analysis to offer improvements in accuracy. We will go ahead and implement ensemble learning methods in building our classification model. To optimize the model, we make use of hyperparameter fine-tuning. In this work, we shall also look at the subtle impact of the learning rates, batch sizes, and other critical parameters to observe what effect it would have on the perfection of the classification model, thereby contributing to a wider understanding of CNNs in medical analysis.