IoT-Driven Predictive Analytics for Smart Medical Devices

The exponentially accelerating growth of intelligent healthcare technologies has necessitated urgent challenges in terms of data privacy, interoperability, and scalability of predictive medical analytics. This research responds to these challenges by implementing an integrated Internet of Things (IoT)-based predictive analytics framework by merging Federated Learning (FL) and Blockchain. The framework facilitates secure, real-time prediction of health risks through the edge processing of physiological data recorded from IoT-integrated medical devices. Locally deployed machine learning algorithms deliver real-time health insights, while FL keeps sensitive patient information decentralized, and Blockchain protects model updates and data transactions using an immutable ledger. Experimental analysis proved that the suggested architecture obtained $98 \%$ predictability accuracy, $98 \%$ privacy, $96 \%$ resilience, and $95 \%$ scalability, outperforming centralized deep learning and traditional machine learning models. The system also minimized decisionmaking latency by $35.5 \%$ and demonstrated robust resilience against adversarial attacks, providing robustness in dynamic healthcare settings. However, shortcomings like infrastructure overhead, heterogeneous device participation, and nonindependence of data distributions still exist. In general, the framework provides a practical and secure basis for scalable predictive healthcare with prospects for application in smart hospitals and remote patient monitoring.