Leveraging IoT-Aware Technologies and AI Techniques for Real-Time Critical Healthcare Applications.

Area of Science:

• * Biomedical Engineering
• * Internet of Things (IoT)
• * Artificial Intelligence (AI)

Background:

• * Wearable devices enhance personalized healthcare by monitoring physiological parameters like heart health and blood glucose.
• * Current cloud-based Machine Learning (ML) approaches for health predictions suffer from latency and privacy concerns.
• * Advanced sensing materials and flexible electronics enable high-accuracy physiological signal measurement.

Purpose of the Study:

• * To review existing IoT healthcare architectures and identify limitations in cloud-dependent ML models.
• * To propose a scalable, modular system architecture leveraging Edge Computing and on-device AI for improved healthcare.
• * To define essential functional and non-functional requirements for critical wearable IoT healthcare systems.

Main Methods:

• * Review of current Internet of Things (IoT) healthcare architectures utilizing wearable devices.
• * Development of a novel system architecture incorporating flexible piezoelectric sensors, low-cost communication, on-device AI, and Edge Computing.
• * Deduction of functional and non-functional requirements based on existing systems and emerging technology trends.

Main Results:

• * A scalable and modular architecture is presented, integrating advanced sensors, on-device intelligence, and Edge Computing.
• * The proposed architecture addresses latency and privacy issues inherent in cloud-based systems.
• * Essential requirements for critical healthcare applications, including modularity, local AI, and data consistency, are outlined.

Conclusions:

• * Edge Computing and on-device AI significantly enhance the efficiency and efficacy of personalized healthcare monitoring.
• * The proposed architecture offers a robust solution for real-time health event prediction and timely intervention.
• * Key requirements for successful wearable IoT healthcare systems include distributed layers, local AI processing, and consistent data packaging.