Privacy-Preserving Vision-Based Detection of Pox Diseases Using Federated Learning

Infectious pox diseases, including monkeypox, chickenpox, and measles, continue to pose significant health challenges globally. Timely and accurate detection is vital for effective disease management and prevention. Traditional diagnostic methods often rely on invasive procedures and may lack privacy safeguards. In response, this research leverages advanced image analysis and federated learning to introduce a privacy-preserving framework for pox disease detection. Our framework achieves an impressive testing accuracy of 87.50% and undergoes a comprehensive evaluation, including key metrics such as Accuracy, Precision, Recall, F1-Score, Loss curve, and Accuracy curve. Furthermore, we explore previous research in vision-based pox disease detection, present our proposed methodology, and discuss experimental results. Emphasizing patient privacy and data security, our study exemplifies the potential of federated learning to revolutionize disease diagnosis while preserving individual confidentiality. This research contributes to enhancing disease management and underscores the significance of privacy-aware healthcare technologies.