Lung Disease Detection Using Resnet and Attention Mechanisms: A Deep Learning Approach

Abstract

The lungs are among the most vital human organs, and their failure can severely disrupt the circulatory and respiratory systems. While numerous studies have sought to detect lung conditions, many focus on the detection of a single disease or rely on disparate algorithms, such as ResNet, DenseNet-101, and ChexNet, without integrating the exploring the possibilities of combining the architectures. This research addresses these limitations by developing a hybrid deep learning model that combines ResNet with an attention mechanism to improve the accuracy of detecting and classifying four distinct lung conditions. To effectively achieve our results, the study will leverage pre-scanned frontal chest X-rays. Our proposed model aspires to serve as a comprehensive diagnostic tool, supporting radiologists in making more informed and accurate decisions.