Wireless IoT and Cyber-Physical System for Health Monitoring Using Honey Badger Optimized Least-Squares...

Area of Science:

• Health Informatics
• Artificial Intelligence in Healthcare
• Cyber-Physical Systems

Background:

• Aging populations necessitate advanced elderly healthcare solutions.
• Internet of Things (IoT) and Cyber-Physical Systems (CPS) are crucial for remote health monitoring.
• Accurate patient health status classification is vital for timely intervention.

Purpose of the Study:

• To propose a novel Honey Badger (HB) algorithm optimized Least-squares Support-Vector Machine (LS-SVM) architecture for elderly health monitoring.
• To enhance the accuracy and efficiency of identifying deteriorating patient health conditions.
• To provide caretakers with timely statistical health reports via a smartphone application.

Main Methods:

• Developed a hybrid model combining the Honey Badger algorithm with LS-SVM for parameter optimization.
• Utilized the HB algorithm to optimize LS-SVM's width coefficient and regularization factor for complex medical datasets.
• Implemented a system for multi-parameter monitoring and classification of patient health status.

Main Results:

• The proposed HB-optimized LS-SVM model achieved high performance metrics: PPV of 0.9478, NPV of 0.9587, and AUC of 0.9617.
• Demonstrated superior accuracy compared to conventional methods like Decision Tree, Random Forest, and SVM.
• Successfully modeled sensor parameters and provided efficient, timely support for elderly patients.

Conclusions:

• The HB-optimized LS-SVM architecture offers a robust and accurate solution for elderly health monitoring.
• The system effectively identifies patients with deteriorating health, enabling proactive care.
• The integration with a smartphone application facilitates seamless data dissemination to caretakers.