An Intelligent Diabetic Patient Tracking System Based on Machine Learning for E-Health Applications

Area of Science:

• Health Informatics
• Artificial Intelligence in Healthcare
• Machine Learning Applications

Background:

• Continuous patient surveillance is crucial for diabetes management.
• Technologies like the Internet of Things (IoT), AI, and 5G networks enable remote and personalized healthcare solutions.
• Increasing healthcare data volume necessitates efficient storage and processing for smart e-health applications.

Purpose of the Study:

• To develop an intelligent system for effective diabetic patient tracking.
• To address challenges in healthcare data management using smart e-health structures.
• To leverage 5G network capabilities for advanced, high-performance healthcare services.

Main Methods:

• An intelligent system for diabetic patient tracking was proposed, utilizing machine learning (ML).
• Data was collected via smartphones, sensors, and smart devices, followed by normalization and feature extraction using Linear Discriminant Analysis (LDA).
• A classification model, Advanced-Spatial-Vector-based Random Forest (ASV-RF) optimized with Particle Swarm Optimization (PSO), was employed for diagnosis.

Main Results:

• The proposed intelligent system demonstrated superior accuracy in diagnostic classification compared to existing methods.
• Simulation outcomes validated the effectiveness of the ASV-RF and PSO integration for diabetic patient tracking.

Conclusions:

• The developed intelligent system offers a promising solution for accurate and efficient diabetic patient monitoring.
• This approach supports the advancement of smart e-health applications and personalized remote healthcare.