Abstract

By using deep learning methods, our study aims to identify pneumonia in chest X-ray pictures using VGG16 convolutional neural networks (CNNs). Our technique ensures data privacy and security by using federated learning to allow collaborative training across distant datasets without centralized data aggregation. We surpass other algorithms, such as ResNet50, in terms of pneumonia diagnosis accuracy. Pneumonia may be easily diagnosed with the use of the Streamlit-based online application. Patients may submit chest X-rays, and the model will indicate whether or not pneumonia is present, along with the likelihood of each diagnosis. By using deep learning with an emphasis on data protection and accessibility, we want to improve healthcare diagnoses via this initiative.