Enhanced Health Information Management with Blockchain-Assisted FL-CNNs and Autoencoders in the Internet of Medical Things

The rapid growth of medical data raises issues in maintaining data security, privacy, and scalability, especially within distributed Internet of Medical Things (IoMT) systems. This study presents a blockchain-enhanced federated learning system including convolutional neural networks and autoencoders for dealing with these challenges. The solution guarantees safe data management using blockchain, improves distributed learning with federated convolutional neural networks, and increases feature extraction and anomaly detection with autoencoders. The results indicate a 98.7% accuracy for federated learning models, a 60% decrease in data dimensionality, and a throughput of 850 transactions per second with a latency of 0.15 seconds. The suggested approach surpasses previous models in accuracy, scalability, and security, rendering it an outstanding solution to modern health information management. These achievements underscore its capacity to transform healthcare through secure, real-time decision-making and thorough patient data analysis.