A Comparative Study of Prediction of Pneumonia and COVID-19 Using Deep Neural Networks

Pneumonia, a contagious respiratory infection caused by both bacteria and viruses, carries significant global implications. Rapid identification of infected individuals is paramount to effectively curb its transmission. This study explores the application of medical image analysis to address this challenge by proposing machine learning techniques for pneumonia prediction using chest X-ray images. Chest X-ray imaging, known for its accessibility and cost-effectiveness, plays a pivotal role in pneumonia diagnosis. However, the interpretation of X-rays for pneumonia detection can be intricate due to potential overlaps with radiographic features of other respiratory conditions. In this research, various machine learning models, including DenseNet121, Inception Resnet-v2, Inception Resnet-v3, Resnet50, and Xception, are evaluated and compared using chest X-ray images from pneumonia patients. Performance metrics and confusion matrices are employed to assess and differentiate these models. The findings highlight DenseNet121 as the top-performing model, achieving an impressive accuracy rate of 99.58%. This study underscores the significance of machine learning in accurately detecting pneumonia through the analysis of chest X-ray images. It offers valuable insights into the potential of technology to mitigate pneumonia's spread by facilitating precise diagnostic tools.