What question did this study set out to answer?

The aim is to develop an adaptive edge-IoT framework using AI to meet the unique health needs of special populations.

February 26, 2026Open Access

AI-agent–empowered edge–IoT framework for personalized health adaptation in special population networks

Key Points

The aim is to develop an adaptive edge-IoT framework using AI to meet the unique health needs of special populations.
Developed the A3E2-IoT framework for real-time health adaptation.
Utilized distributed smart agents for data processing and decision-making.
Employed a hierarchical collaboration system encompassing edge, fog, and cloud levels.
Implemented an adaptive learning function with federated intelligence.
Reduced system latency by 42% compared to traditional cloud-based models.
Enhanced anomaly detection accuracy by 15%.
Decreased cloud bandwidth consumption by 37%.
Improved accuracy by 7.8%, precision by 6.1%, recall by 6.3%, and F1-Score by 6.8%.
Increased adaptability by 10.2% and privacy compliance by 6.0%.

Abstract

The integration of Artificial Intelligence (AI) and the Internet of Things (IoT) has brought about groundbreaking transformations in intelligent healthcare, enabling decisions based on data, early detection of diseases, and monitoring of patients around the clock. However, some populations, such as the elderly, disabled, and those with chronic illnesses, require health management that is tailored to their unique needs and circumstances. Issues such as latency, low bandwidth, and privacy issues indicate that existing cloud-based models are not always suitable for real-time adaptive treatment. To overcome such issues, this research demonstrates an A3E2-IoT framework, which refers to AI-Agent-Adaptive Empowered Edge-IoT. This framework is designed to address the distinct health requirements of various groups of individuals. The architecture employs smart agents that are distributed throughout the edge of the network to deal with data processing, decision-making, and modification. These agents utilize AI models that are not overly resource-intensive. To ensure health monitoring, anomaly detection, and personalized recommendations occur in real time, a hierarchical collaboration system is employed at the edge, fog, and cloud levels. The adaptive learning function leverages federated intelligence to continue refining local models while maintaining user privacy and accuracy. Experimental verification across simulated multi-patient datasets reveals that A3E2-IoT decreases system latency by 42%, enhances anomaly detection accuracy by 15%, and reduces cloud bandwidth consumption by 37% over conventional IoT systems using the cloud. Comprehensive experiments across key parameters such as accuracy (+ 7.8%), precision (+ 6.1%), recall (+ 6.3%), F1-Score (+ 6.8%), latency (− 21.5%), aggregation time (− 24.3%), adaptability (+ 10.2%), and privacy compliance (+ 6.0%). Finally, the A3E2-IoT framework establishes a framework that is intelligent, flexible, and ensures privacy protection for intelligent health management. By integrating distributed AI agents with enabled edge computing, this emerging technology enables the possibility of delivering next-generation individualized healthcare to targeted population networks.

Bookmark

View Full Paper