Fused federated learning framework for secure and decentralized patient monitoring in healthcare 5.0 using IoMT.

Federated Learning (FL) enables artificial intelligence frameworks to train on private information without compromising privacy, which is especially useful in the medical and healthcare industries where the knowledge or data at hand is never enough. It paved the way for a substantial amount of study because of the high degree of communication efficacy it possessed, which is connected to dispersed training issues. The major goal of this paper is to shed light on how FL approaches might be adapted and put to use in several aspects of healthcare, including medicine discovery, medical assessment, digital health management, and the forecasting and identification of disease. This article presents a comprehensive and in-depth study of the data about fused federated learning in healthcare version 5.0. The purpose of this research is to develop a Healthcare 5.0 monitoring system by utilizing a fused federated learning approach integrated with RTS-DELM. It gives medical practitioners the ability to monitor patients through the use of various medical sensors and to take remedial action at regular intervals. The approach is shown to be successfully improved by the use of the recommended system, which is intended for healthcare monitoring. This paper introduces a novel framework leveraging Fused Federated Learning (FFL) integrated with IoMT devices aimed at securely monitoring patient health data in a decentralized manner. This study introduces a novel integration of Real-Time Sequential Deep Extreme Learning Machine (RTS-DELM) and Fused Federated Learning (FFL) for secure and decentralized chronic kidney disease diagnosis within Healthcare 5.0. The proposed approach efficiently aggregates data from distributed Internet of Medical Things (IoMT) devices, enhancing predictive accuracy while maintaining patient privacy. Experimental validation demonstrates significant improvements, achieving an accuracy rate of 98.21%, thereby showcasing superior performance over existing federated learning methodologies.